Evolving Men IMAGES

Evolving Men
Initial Results from the International Men and Gender Equality Survey
(IMAGES)
coordinated by the International Center for Research on Women and Instituto Promundo
About the Men and Gender Equality Policy Project

The Men and Gender Equality Policy Project (MGEPP), led by Instituto Promundo and the International Center for Research on Women (ICRW), is a multi-year, multi-country effort to build the evidence base on how to change public institutions and policies to better foster gender equality and to raise awareness among policymakers and program planners of the need to involve men in health, development and gender equality issues. Project activities include: (1) a multi-country policy research and analysis presented in the publication What Men Have to Do with it: Public Policies to Promote Gender Equality; (2) the International Men and Gender Equality Survey or IMAGES, a quantitative household survey carried out with men and women in seven countries in 2009-2010,1 initial results of which are presented in this publication; (3) the Men who Care study consisting of in-depth qualitative life history interviews with men in five countries; and (4) advocacy efforts and dissemination of the findings from these different components via various formats, including a documentary film. Participating countries in the project as of 2010 included Brazil, Chile, Croatia, India, Mexico, Rwanda and South Africa. The multiple research components of the project aim to provide policymakers with evidence-based, practical strategies for engaging men in gender equality, particularly in the areas of sexual and reproductive health, reducing gender-based violence, fatherhood and maternal and child health, and mens health needs.
Photo Credits cover (top to bottom): David Isaksson/Influential Men David Herdies/Influential Men Pernilla Norstrm/Influential Men David Snyder/ICRW facing page, all photos: David Snyder/ICRW Design by Barbieri and Green
Evolving Men
Initial Results from the International Men and Gender Equality Survey
(IMAGES)
Men and Gender Equality Policy Project Coordinated by the International Center for Research on Women and Instituto Promundo
Evolving Men Initial Results from the International Men and Gender Equality Survey (IMAGES)
Acknowledgments
We extend our special thanks for contributions to the IMAGES questionnaire and the research process to: Rachel Jewkes, Medical Research Council, South Africa; Margaret Greene, GreeneWorks Global; Jorgen Lorentzen, Center for Gender Research, University of Oslo, Norway; Oystein Gullvg Holter, Center for / a Gender Research, University of Oslo; Robert Morrell, University of Cape Town, South Africa; Abhijit Das, Centre for Health and Social Justice, India; Dean Peacock, Sonke Gender Justice Network, South Africa; Benedito Medrado, Instituto Papai, Brazil; and Elizabeth Sussekind, Pontifcia Universidade Catolica, Rio de Janeiro, Brazil. We also thank James Lang and Emma Falu of Partners for Prevention for their input and support and Ram Manohar Mishra for his contriGary Barker, Juan Manuel Contreras, Brian Heilman, butions to the cleaning and merging of the datasets. Ajay Singh and Ravi Verma, all of the International We are grateful to a number of ICRW colleagues who Center for Research on Women, and Marcos Nascicontributed to the report and the research process, mento of Instituto Promundo coordinated the including Ellen Weiss, Sandy Won, Lindsay Kin and International Men and Gender Equality Survey Mary Ellsberg. At Promundo, we would like to thank (IMAGES), carried out this initial comparative Fabio Verani, Hugo Correia, Rafael Machado, Anna analysis and authored this report. Luiza Almeida and Piotr Pawlak. Special thanks as well to John Crownover, CARE NW-Balkans. We thank The country surveys were coordinated and authored by: Viktoria Saxby Roxie of Hope Exhibits as well as the photographers whose images from "Influential Men," Brazil: Mrcio Segundo, Marcos Nascimento and a global photography exhibit, are included in this Christine Ricardo, Promundo report. Finally, we extend our gratitude to all the Chile: Francisco Aguayo and Michele Sadler, interviewers and to all the women and men who CulturaSalud agreed to be interviewed for the study. Croatia: Natasa Bijelic, CESI-Center for Education, Counseling and Research India: Ravi Verma and Ajay Singh, International Center for Research on Women Mexico: Juan Guillermo Figueroa, El Colegio de Mxico, Mxico, D.F. Rwanda: Henny Slegh, Augustin Kimonyo, Fidle Rutayisire and Edouard Munyamaliza, Rwandan Mens Resource Centre (Rwamrec) This publication, the data collection and the data analysis were generously funded by the John D. and Catherine T. MacArthur Foundation, the Government of Norway (Ministry of Foreign Affairs and Norad), the Ford Foundation and an anonymous donor. Portions of the analysis and country-level work were supported by CARE-Norway (Croatia); UNDP (Rwanda); and UNIFEM and UNFPA (Chile). This publication was produced as part of the Men and Gender Equality Policy Project, coordinated by Instituto Promundo and the International Center for Research on Women (ICRW). All conclusions and opinions are those of the authors and do not reflect the official position of any of the donors.
2011 International Center for Research on Women (ICRW) and Instituto Promundo. Portions of this report may be reproduced without express permission from but with acknowledgment to ICRW and Instituto Promundo. Suggested citation: Barker, G., Contreras, J.M., Heilman, B., Singh, A.K., Verma, R.K., and Nascimento, M. Evolving Men: Initial Results from the International Men and Gender Equality Survey (IMAGES). Washington, D.C.: International Center for Research on Women (ICRW) and Rio de Janeiro: Instituto Promundo. January 2011.
TABLE OF CONTENTS
Acronyms Summary of Findings PART I: Why the International Men and Gender Equality Survey (IMAGES)?
Introduction, Methodology and Limitations Conceptual Framework, Topics and Analytical Strategy
6 7 10
10 14
PART II: Results

Who Are the Women and Men who Participated in IMAGES? Mens Work-Related Stress Gender Attitudes Mens Practices and Lives Relationship Dynamics and Domestic Duties Parenting and Involvement in Childbirth Health Practices and Vulnerabilities Violence and Criminal Practices Transactional Sex Knowledge and Attitudes about Policies and Laws Related to Gender Equality
16
16 16 18 22 22 28 35 43 49 55
PART III: Conclusions References Annexes

Annex I: Overview of Related Studies Annex II: Original Conceptual Framework Annex III: GEM Scale Scores Trichotomized by Country Annex IV: Sampling Methodologies Annex V: Intimate Partner Violence Prevalence, IMAGES and other Recent Studies Annex VI: Tests of Association
60 62
64 68 69 70 75 76
Notes
98
Tables and Figures

Tables
1 2 3 4 IMAGES Data Collection Details Socio-Demographic Characteristics, Men Socio-Demographic Characteristics, Women Percent of Men who Agree or Partially Agree with GEM Scale Items by Country 5 Mens Reports of Equal or Greater Participation in Domestic Duties, in Order from Most to Least Common Overall
13 16 17 19
6 Mens Reports of Physical Violence Against a Female Partner 7 Womens Reports of Physical Violence by an Intimate Male Partner 8 Mens Reports of Criminal Activity, Firearm Ownership and Imprisonment 9 Men's Reports of Transactional Sex
43 43 47 49
24
Marie Swartz/Influential Men
TABLE OF CONTENTS
Figures
1 Mens Reports of Work-Related Stress 2 Mens Attitudes about Homosexuality 3 Mens and Womens Reports about Speaking with Partner about a Personal Problem in Previous Month
18 20
19 Men Older Than Age 40 Who Have Ever Sought A Prostate Exam 36 20 Mens and Womens Reports of Having Ever Sought an HIV Test
22
21 Reports of Induced Abortion 22 Mens and Womens Reports of Mens Involvement in Abortion Decision-Making (of Those Who Report That They or a Partner Ever Sought an Abortion) 23 Mens Affirmations of Positive Self-Esteem 24 Mens Negative Statements of Well-Being 25 Mens Reports of Feeling Depressed Sometimes or Often in the Last Month 26 Mens Reports of Suicidal Thoughts (Sometimes or often Thinking about Suicide in the Past Month) 27 Association Between Mens Reports of Work-Related Stress and Suicidal Thoughts
37 38
4 Mens and Womens Reports of Mens Participation in Domestic Duties: Percent Who Report That Men Play an Equal or Greater Role in Household Duties (Excluding Child Care) 23 5 Links Between Fathers and Sons Participation in Domestic Duties (Defined as Playing an Equal or Greater Role in One or More Duties) 24 6 Mens and Womens Reports of Sexual Satisfaction With their Current Partner 7 Mens Reports of Sexual Satisfaction and their Reports of Couple Communication 8 Womens Reports of Sexual Satisfaction and their Reports of Couple Communication 9 Mens and Womens Reports of Satisfaction with Current Division of Household Duties 10 Mens Presence during Birth of Last Child 11 Mens Presence in the Delivery Room by Age in Chile 12 Mens and Womens Reports of Mens Accompaniment during Prenatal Visits during Most Recent Pregnancy
38 39 40 40
25 26 27 27 28 29
41 41
28 Mens and Womens Reports of Seeking Help When they Feel Sad, Disappointed or Frustrated 42 29 Mens Reports of Using IPV Against Female Partner and Womens Reports of Experiencing IPV from A Male Partner (Lifetime) 30 Mens Reports of Using Sexual Violence 31 Transactional Sex Experiences: Mens Reports That a Sex Worker With Whom they Had Sex Was Under 18 or Forced/Sold into Prostitution
44 46
30
13 Parental Leave after Birth of Most Recent Child Among Currently Working Men 31 14 Mens and Womens Reports of Mens Participation in the Daily Care of A Child 15 Mens Reports of Care of Children Age 4 and Younger 16 Womens Reports of Mens Participation in Care of Children Age 4 and Younger 17 Mens Employment Status and Daily Care of a Child 18 Mens Reports of Regular Abuse of Alcohol Compared to Womens Reports of (their own) Regular Abuse of Alcohol (Five or More Drinks on One Occasion, Once Monthly or More)
50 32 Mens Attitudes about Adult Female Sex Workers 51 33 Mens Attitudes about Adult Male Sex Workers 52
34 Mens Attitudes about Female Sex Workers Younger Than 18 36 Men's Support for Gender Equality Overall and Specific Quota-Based Policies 37 Mens Exposure to Messages and Campaigns about Violence Against Women 38 Knowledge about VAW Law, Attitudes about VAW, and Perpetration of Violence 39 Mens Attitudes about Laws Criminalizing Intimate Partner Violence
32 32 33 34
53 35 Mens Attitudes about Male Sex Workers Under 18 54 55 56 57 58
35
40 Mens Exposure to Messages and Campaigns Related to Fatherhood and Related to Sexual Diversity 59
Acronyms
CSW DHS GBV GEM ICPD IMAGES IPV MDGs MGEPP MRC PPS SRH STIs VAW WHO Commission on the Status of Women Demographic and Health Survey Gender-Based Violence Gender-Equitable Men (Scale) International Conference on Population and Development International Men and Gender Equality Survey Intimate Partner Violence Millennium Development Goals The Men and Gender Equality Policy Project Medical Research Council of South Africa Probability Proportion to Size Sexual and Reproductive Health Sexually Transmitted Infections Violence Against Women World Health Organization
SUMMARY OF FINDINGS
Summary of Findings
About IMAGES
The International Men and Gender Equality Survey (IMAGES) is a comprehensive household questionnaire on mens attitudes and practices along with womens opinions and reports of mens practices on a wide variety of topics related to gender equality. From 2009 to 2010, household surveys were administered to more than 8,000 men and 3,500 women ages 18-59 in Brazil, Chile, Croatia, India, Mexico and Rwanda. Topics in the questionnaire included: gender-based violence; health and healthrelated practices; household division of labor; mens participation in caregiving and as fathers; mens and womens attitudes about gender and genderrelated policies; transactional sex; mens reports of criminal behavior; and quality of life. This report focuses on the initial, comparative analysis of results from the mens questionnaires across the six countries, with womens reports on key variables. more than 80 percent in India agreed with the statement. Men with higher educational attainment and married men had more equitable attitudes; unmarried men had the least equitable attitudes. Homophobic attitudes were common although varied tremendously by context. Men who said they would be ashamed to have a gay son ranged from 43 percent of men in Brazil to a high of 92 percent in India. A slightly lower, but still high proportion of men said that being around homosexual men makes them uncomfortable, ranging from a low of 21 percent of men in Brazil to a high of 89 percent in India. Younger men and men with higher levels of education were generally less homophobic.
Relationship Dynamics and Domestic Duties Younger men, men with more education and men who saw their fathers do domestic work are more likely to carry out domestic duties. Nearly half of men in all the sites say they play an equal or greater Key Findings role in one or more household duties with the exception of India, where only 16 percent of men Work-Related Stress reported that they played an equal or greater role in Work-related stress is commonplace in all survey household duties. These household or domestic sites. Between 34 percent and 88 percent of men in the survey sites reported feeling stress or depression duties included washing clothes, repairing the house, buying food, cleaning the house, cleaning the because of not having enough income or enough bathroom or toilet, preparing food and paying the work. Men who experienced work-related stress were more likely to report depression, suicide ideation, previous arrests, and use of violence against intimate partners. Gender Attitudes Men showed tremendous variation in their genderrelated attitudes, with India and Rwanda showing the most inequitable attitudes. As a measure of mens and womens gender-related attitudes, IMAGES applied the Gender-Equitable Men (GEM) Scale.2 Rwandan and Indian men consistently supported the least equitable norms among the settings studied. For example, for the statement changing diapers, giving kids a bath and feeding kids are the mothers responsibility, only 10 percent of men agreed in Brazil whereas 61 percent in Rwanda and
Jon Spaull/Influential Men
Health Practices and Vulnerabilities Mens rates of regular abuse of alcohol defined as having five or more drinks in one night on a once monthly or greater basis vary from 23 percent in India to 69 percent in Brazil and are significantly Men report higher levels of sexual and relationship higher than womens reported alcohol abuse in all satisfaction than women. Women who say their survey sites. In most sites, younger men and men partners do more domestic work are more sexually with more inequitable gender attitudes are more satisfied. Men reported relatively high rates of sexual likely to regularly abuse alcohol. satisfaction with their current stable partners ranging from 77 percent in Croatia to 98 percent High proportions of women who reported having in India. In all the countries except India, men who sought an abortion affirmed that a male partner was reported more gender-equitable attitudes were more involved in the decision to seek an abortion (ranging likely to report being sexually satisfied with their from 39 percent to 92 percent). current female partner. In India, Brazil and Croatia, women who reported that their male partner plays Men report high self-esteem, with the exceptions of an equal or greater role in one or more domestic Croatia and India; at the same time, men show duties also reported higher levels of overall relation- relatively high levels of depression and suicide ideation. ship satisfaction and sexual satisfaction. The rates of experiencing depression at least once in the last month ranged from 9 percent in Brazil to a Parenting and Involvement in Childbirth high of 33 percent in Croatia. The percentages of male The majority of men were neither in the delivery respondents who reported having suicidal thoughts room nor the hospital for the birth of their last child. "sometimes or often" in the last month ranged from In Chile, however, a dramatic generational shift is 1 percent in Brazil and Mexico to 5 percent in Croatia. underway in mens presence at childbirth. Younger Violence and Criminal Practices Chilean men reported much greater rates of presence in the delivery room for the birth of their last Men reported lifetime rates of physical intimate partchild than older men. This shift is largely due to a ner violence ranging from 25 percent to 40 percent national policy, aimed at humanizing the birth with women reporting slightly higher rates. Factors process, which encourages women to have a male associated with mens use of violence were rigid partner or other person of their choice present dur- gender attitudes, work stress, experiences of violence ing birth at public maternity wards. in childhood and alcohol use. Mens reports of perpetration of sexual violence against women and girls Men are taking few days of paid or unpaid paternity ranged from 6 percent to 29 percent; in India and leave. Among men who took leave, the duration Mexico, the majority of sexual violence took place ranged from an average of 3.36 to 11.49 days of paid against a current or former partner. leave and from an average of 3.8 to 10 days unpaid leave. Younger men and men with more education Relatively high percentages of men reported ever havwere more likely to take leave. ing participated in criminal or delinquent acts; between 6 percent and 29 percent of men report ever Close to half of men with children say they are involved having been arrested. In terms of factors associated in some daily caregiving. Unemployed men are dramat- with mens participation in criminal activity, mens ically more likely to participate in the care of children socio-economic situation was the most significant. than employed men. For men with children under age Men who owned firearms or carried out other viofour, play is the most common daily activity in which lence or criminal behavior were also more likely to they participate (as affirmed by women and men). report having used intimate partner violence.
bills. The tasks that men said they play an equal or greater role in are those traditionally associated with men namely repairing the house, paying bills and buying groceries.
SUMMARY OF FINDINGS
Transactional Sex Between 16 percent and 56 percent of men in the sites surveyed say they have paid for sex at least once. Men with lower educational attainment, less gender-equitable attitudes and men who reported less sexual satisfaction with their current partner are more likely to have paid for sex.
Conclusion
Overall, IMAGES results affirm that gender equality should be promoted as a gain for women and men. Change seems to be happening as younger men and men with higher levels of education show more gender-equitable attitudes and practices. Men who report more gender-equitable attitudes are more Knowledge and Attitudes about Policies and Laws likely to be happy, to talk to their partners and to have better sex lives. Women who report that their Related to Gender Equality partners participate in daily care work report higher Men in all the countries, with the exception of India, levels of relationship and sexual satisfaction. Findare generally supportive of gender equality, with 87 percent to 90 percent saying that men do not lose out ings suggest that most men in most of the survey when womens rights are promoted. Even when asked sites accept gender equality in the abstract even if about specific policies quotas for women in execu- they are not yet living it in their daily practices. tive positions, in university enrollment or in government mens support for such policies is reasonably high, with 40 percent to 74 percent of men supporting such quotas. Among themes related to gender equality, men reported the highest exposure to campaigns about gender-based violence. At the same time, across the sites, men showed negative attitudes toward laws related to gender-based violence.
Pernilla Norstrm/Influential Men
PART I: Why The International Men and Gender Equality Survey (IMAGES)?
Introduction, Methodology and Limitations
Background The world increasingly affirms that men must be part of achieving gender equality. Indeed, changing mens practices and the structures and factors that enable, encourage and shape those practices in terms of violence, health, overall treatment of women and girls and participation in family life is a key part of the global gender equality agenda. The United Nations has called for engaging men and boys in gender equality for at least 15 years. The 1994 International Conference on Population and Development (ICPD, Cairo) affirmed the importance of involving men in improving sexual and reproductive health, and emphasized the need to increase mens involvement in the care of children (WHO, 2007). The ICPD Programme of Action calls for leaders to promote the full involvement of men in family life and the full integration of women in community life, ensuring that men and women are equal partners. age men to fully participate in the care and support of others, particularly children.4 Many of the Millennium Development Goals (MDGs), which are time-bound targets for achieving key health and development indicators, focus on achieving gender equality and improving womens lives (in terms of maternal health, educational disparities, poverty eradication, HIV/AIDS prevention and care, and reducing violence against women). An analysis using 12 indicators to measure progress on the MDG priorities (ICRW, 2008) shows that clear advances have been made in empowering women, especially in education and increasing womens political representation in national parliaments and local governments. But progress in other areas that require engaging men reducing violence against women, increasing womens income relative to mens, and reducing inequalities related to the care burden has lagged.
Ping-hang Chen/Influential Men
The question is: are men on board with the gender equality agenda? Are they internalizing the messages and policies calling for greater equality for Similarly, the 48th session of the Commission on the girls and women in education, income and work, Status of Women3 (CSW) in 2004 affirmed the ICPD political participation and health? Are men changand went further, recognizing that men and boys ing their attitudes, practices and the ways they live can and do make contributions to achieve gender their lives in relation to women? Are men changing equality. CSW urged governments to adopt and practices related to their own health and interperimplement policies to close the gap between women sonal relationships? How much are men and men in terms of occupational segregation, participating in the care of children and other parental leave and working arrangements to encour- domestic activities? How common is mens use of violence against intimate partners? What factors are associated with this violence? What do men think about existing laws on gender-based violence and other policies designed to promote gender equality? How do these issues differ in various regions of the world? And, are mens own lives improving as they embrace gender equality and take on more equitable, flexible and non-violent versions of masculinity? In other words, are men evolving to be more supportive of gender equality? While many of these questions have been asked in numerous individual studies, they have seldom been
10
PART 1
PART 2
PART 3
asked together as part of a multi-country initiative This report provides an initial, comparative analysis of: using statistically representative household samples. Mens socio-demographic status, including employment status and employment-related stress The Men and Gender Equality Policy Project Mens self-reported attitudes and practices, (MGEPP) is a multi-year, multi-country effort to along with womens reports of mens behaviors gather evidence and raise awareness among policyon some of these practices, related to relationship makers and program planners of the need to involve dynamics, parenting and involvement in men in gendered ways as part of health and developchildbirth, health practices, violence (including ment agendas. A key component of the project has intimate partner violence), and transactional sex been the International Men and Gender Equality Mens reported knowledge and attitudes toward Survey (IMAGES), one of the most comprehensive existing gender equality policies efforts of its kind to gather household survey data on mens attitudes and practices along with womens This report provides descriptive statistics and biopinions about and experiences of mens practices variate analyses of key results for the topics listed on a wide variety of topics related to gender equality. above. Individual country reports will provide more analyses of country results. Future reports will anaThis report presents initial findings on mens lyze results in more depth on specific topics as well attitudes and practices from six countries where the provide multi-variate analyses of key variables. survey was first carried out: Brazil, Chile, Croatia, India, Mexico and Rwanda. These countries represent Methodology IMAGES followed standard procedures for carrying different geographic regions where advocacy efforts related to engaging men in gender equality, many of out representative household surveys in each particithose in collaboration with MenEngage or other pating city, with the exception of Rwanda, where the networks and civil society groups, are underway.5,6 survey is a nationally representative household sample. As seen in Table 1, the survey was carried out in About the International Men and Gender Equal- one or more urban settings in each country (and rural ity Survey (IMAGES) and urban areas in Rwanda) with men and women The overall goal of IMAGES is to build understandages 18-59, guided by the following parameters: ing of mens practices and attitudes related to gender equality in order to inform, drive and monitor policy The mens questionnaire has approximately 250 development to promote gender equality by engagitems and took from 45 minutes to an hour to ing men and women in such policies. IMAGES seeks apply; the questionnaire for women is slightly to assess the current practices and attitudes of men shorter and took from 35 minutes to an hour to on a range of issues related to gender equality, administer. The survey instruments were preincluding fatherhood and caregiving (time spent in tested in the participating countries and the study providing care for children, and others in the home), protocol was approved by ICRWs institutional use of violence in intimate and sexual relationships review board (IRB) and by in-country IRBs, when (sexual, physical and psychological) and against such existed. other men, work-life balance, use of health services, The survey instrument was designed to be relevant negotiation/communication with partner about for adult men and women in stable, co-habitating family size and sexual relations and existing gender relationships as well as those not in a stable relationequality policies, among others. IMAGES incorpoship; women and men who define themselves as rates items from the most recent survey instruments heterosexual as well as men and women of different on gender, quality of life, childhood antecedents of sexual orientations and practices; and women and violence (including gender-based violence), health, men who have children in the household (biological family gendered dynamics and fatherhood.7 or otherwise) and those who do not.
11
Double-back translation of the questionnaire was carried out to ensure comparability and consistency of questions across settings. Some country-specific questions were included; some countries excluded items due to local political or cultural considerations. In Brazil, Chile, Mexico and Rwanda, the questionnaire was an interviewer-administered paper questionnaire. In India, the questionnaire was carried out using hand-held computers, with a mixture of self-administered questions and interviewer-asked questions. In Croatia, the questionnaire was self-administered (using a paper questionnaire). Standard procedures were followed for ensuring anonymity and confidentiality. All research sites followed standard WHO practices for carrying out research on intimate partner violence in terms of offering referrals and information for services and special training of interviewers. 8 Following these guidelines, men and women were not interviewed from the same household in any of the research sites. More sensitive questions were asked later in the questionnaire, and some key variables were included in multiple questions (to compare and thus be more informed in affirming validity). The questionnaire was pre-tested in all the settings prior to application. In all settings, male interviewers interviewed male respondents and female interviews interviewed female respondents, with the exception of Mexico, where some interviews with men were carried out by female interviewers (but only women interviewed women). Survey locations were chosen to represent different contexts in each country to achieve a mixture of major urban areas and a secondary city or cities. Within a survey location, neighborhoods or blocks were chosen based on population distributions from the most recent census data. Rural areas were included only in Rwanda and Croatia. Stratified random sampling and probability proportion to size (PPS) sampling methods were used within each neighborhood or community to ensure the inclusion of adequate sample sizes by age and residence (and also socio-economic status in the case of Chile).9
Although all participating countries included questions on all the themes that make up IMAGES, the questionnaire is not identical in all countries, thus data is not available from every country for every question. The questionnaire in Rwanda was the most abbreviated of the six study countries, due to the much larger sample size and thus the sheer number of interviews required to make the study nationally-representative. In those cases where Rwandan data does not appear in a table or figure in this document, that particular question was not included in the Rwandan questionnaire. Limitations IMAGES was carried out as a city-based, random household survey. The findings presented here are representative of individual cities where the survey was carried out and not of their countries as a whole (except in the case of Rwanda, where the data are nationally representative). Throughout this report, city data are aggregated in order to present overall percentages for each country, but the initial results presented here are, strictly speaking, only representative of their city or neighborhood settings. Individual country reports and future multi-variate analyses will provide separate results by city. Other limitations to IMAGES include the length of time it took for an individual to respond to the entire questionnaire, which led to fatigue on the part of interviewers and respondents in some contexts, as well as the challenge of getting middle-class households to respond in most settings. The fact that female interviewers carried out interviews with men in some cases in Mexico may have also produced some biases in the case of data from Mexico. Other factors also likely contributed to refusal rates or may have biased results, particularly ongoing urban violence in the cases of Brazil and Mexico. Interviewers in research sites in those two countries noted high levels of tension and reluctance on the part of respondents to allow interviewers to enter their homes. Interviewers suspect that the ongoing violence in both settings may have biased some of mens reports on the use of different forms of violence.
12
PART 1
Related Studies Many items on the IMAGES questionnaire have been influenced by a survey designed for the 2005 Gender Equality and Quality of Life study in Norway, carried out by the Nordic Gender Institute (NIKK) and the Work Research Institute (WRI) and financed by the Norwegian Ministry of Children and Equality.10 A second study on men, health and violence, carried out in 2008-2009 by the Medical Research Council of South Africa, used many but not all items from the IMAGES questionnaire and
PART 2
PART 3
added many items for the IMAGES questionnaire.11 Brief overviews of both of these studies can be found in Annex I. Experts at ICRW, Promundo and the research partner organizations encourage investigators in other countries to apply IMAGES and are available to provide further information and/or technical assistance. Similarly, researchers interested in carrying out additional analyses from the IMAGES datasets are encouraged to contact either Promundo or ICRW.
Data Collection Details Sample size, men Sample size, women Age group Site(s)
Brazil
750 448 18-59 One major metropolitan area: Rio de Janeiro, with two neighborhoods: Mar (low income) (n=686 men, 408 women) and Vila Valquiere (middle income) (n= 64 men, 40 women) Promundo
Chile
1192 426 18-59 Three metropolitan areas: Valparaso (n=198 men), Concepcin (n=197 men), Santiago (n=797 men, 426 women)
Croatia
1501 506 18-59 One metropolitan area and two rural areas: Zagreb (n=809 men, 264 women), and towns and villages in two counties in Eastern Croatia (n=692 men, 242 women)
India
1534 521 18-59 Two metropolitan areas: Delhi (n=1037 men, 313 women) and Vijayawada (state of Tamil Nadu) (n=497 men, 208 women)
Mexico
1001 383 18-59 Three metropolitan areas: Monterrey (n=515 men, 172 women); Queretaro (n=222 men,127 women); Jalapa (n=264 men, 84 women)
Rwanda
2301 1311 18-59 Nationally representative sample covering all provinces: Eastern province (25%), Kigali (11%), Northern province (19%), Southern province (25%), Western province (20%)
In-country research partner
CulturaSalud, EME CESI-Center for Education, Counseling and Research Paper Survey (intervieweradministered) Paper Survey (mix of self administered and interviewer administered)
ICRW, SIGMA Research and Consultancy Pvt. Ltd. Handheld Survey (mix of self administered and interviewer administered)
El Colegio de Mxico, Rwanda Mens Mxico, D.F Resource Centre (RWAMREC)
Questionnaire Application Process
Paper Survey (mix of self administered and interviewer administered)
Paper Survey (mix of self administered and interviewer administered)
Paper Survey (interviewer administered)
Table 1 - IMAGES Data Collection Details
13
Conceptual Framework, Topics and Analytical Strategy

Conceptual Framework Gender refers to the widely shared expectations and norms within a society about appropriate male and female roles, responsibilities and behaviors, and the ways in which women and men interact with each other (Gupta, 2000). Gender therefore encompasses masculinities and femininities, power relations between women and men, and structural contexts that reinforce and create these power relations. This study is framed within this understanding of gender as relational and structural, and within the field of masculinities, which seeks to understand how men are socialized, how mens roles are socially constructed (in constant interaction with womens roles), and how these roles and power dynamics change over the lifecycle and in different social contexts (Connell, 1994). This concept of masculinities also enjoins us to examine the diversity of men, the pressures they may perceive to adhere to specific versions of manhood particularly the widespread belief that being a man means being a provider or having stable employment and to understand how these roles and power dynamics change historically and by social context, and within the various domains where social meanings of gender and gender-based inequalities are constructed. In keeping with a relational perspective, a similar instrument was applied to both women and men, with an understanding that mens and womens understanding and reports of their respective practices may be different. By including and comparing responses from women and men, we do not take the approach that either mens or womens responses are necessarily more valid than the other. Where there are significant differences in womens and mens reports of a given practice, we reflect on why this may be the case. Topics The questionnaire covers key topics in gender equality, including intimate relationships, family dynamics, and key health and social vulnerabilities
for men. Based on previous research that found associations between early childhood exposure to violence and different gendered practices related to childrearing, we also included items on childhood antecedents to particular mens practices. Specific topics in the questionnaire include:12 1. Employment. Employment experience; unemployment and underemployment; stress and reactions associated with unemployment; reaction by spouse/partner when unemployed; income differentials between men and women; perceived gender dynamics in the workplace; work-life balance; and job satisfaction. 2. Education. Educational attainment; perceived gender norms and patterns in school. 3. Childhood experiences. Victimization by violence as children; witnessing of gender-based violence; gender-related attitudes perceived in family of origin; changes perceived from previous generation to the present; gender balance in work/child care in family of origin; genderpatterns of childhood friendships. 4. Relations at home (in current household). Marital/cohabitation status; division/participation in household chores; perceived satisfaction in family life; household decision-making; time use in specific domestic chores and family care, including child care. 5. Parenting and mens relationships with their children (and with non-related children who may live in the household). Number of children; living situation of each child; time/money spent in care of each child; use of paternity/maternity leave; perceptions/attitudes toward existing parental leave in country; and child care arrangements. 6. Attitudes toward women and masculinity. Attitudes toward gender equality (using the GEM Scale and other measures); attitudes toward various gender equality policies that may have been implemented in each country. 7. Health and quality of life. Lifestyle questions (substance use, exercise, etc.); use of health services; sexual and reproductive behavior (contraceptive use, condom use); sexually transmitted infections including HIV (past history,
14
PART 1
HIV testing); satisfaction with sexual relations; mental health issues (depression, suicide ideation); social support; use of/victimization of violence in other contexts; morbidity.13 8. Partner relations and spousal relations. Current relationship status/satisfaction; use of services/help-seeking in times of violence or relationship stress; relationship history. 9. Relationship, gender-based violence and transactional sex. Use of violence (physical, sexual, psychological) against partner (using WHO protocol); victimization of violence by partner (using WHO protocol); mens use of sexual violence against non-partners; mens selfreported purchasing of sex or paying for sex, including with underage individuals. 10. Sexual behavior. Sexual experience; sexual orientation; behaviors related to sexual and reproductive health, HIV/AIDS; use of health services related to sexual and reproductive health.
PART 2
PART 3
Analytical Strategy This report focuses on mens attitudes and practices related to relationship dynamics, parenting and caregiving, health-related practices and vulnerabilities, violence (intimate partner violence and other forms), transactional sex and attitudes toward existing gender equality policies. Womens reports of mens practices are reported for some key variables. Future reports will provide further analyses of womens results. The selection of questions in this initial data analysis was informed by previous research confirming the associations or impact of early childhood experiences, individually held gender-related attitudes, educational attainment, age (as a proxy of generational differences as well as developmental stage), social class (or income) and employment status and economic stress on womens and mens attitudes and practices in terms of their intimate relationships, their sexual practices, their use of violence, their domestic practices and their healthseeking behaviors. The framework used for the data analysis is included as Annex II. Mens knowledge of and attitudes toward key policy issues related to gender equality are also included. This report focuses on descriptive statistics and bi-variate analyses of the associations between educational levels, economic or work-related stress, genderrelated attitudes, and age and these practices. In all cases where statistically significant differences are reported, these are at the p < .05 level as assessed using the Pearsons chi-square test. As noted earlier, we also have an interest in understanding generational changes, or changes over time, in terms of mens practices. IMAGES is not a longitudinal study; nonetheless, by comparing responses stratified by age groups we can make some inferences about generational change.
David Isaksson/Influential Men
15
PART II: Results

Who Are the Men and Women who Participated in IMAGES?
Table 2 provides a general overview of the men and women interviewed, in terms of age, marital status, educational levels and parenting status. As would be expected given the focus on adult men ages 18-59, the majority of men and women interviewed in all sites are married with children. In general, men have completed primary education with some secondary education. The samples are reasonably comparable to general demographics in urban areas in their respective countries. Mens Work-Related Stress In addition to asking men about employment and household income (the usual variables for assessing
Brazil n=750
% N
Chile n=1192
% N
Croatia n=1501
% N
India n=1534
% N
Mexico n=1001
% N
Rwanda n=2301
% N
Age
18-24 25-34 35-49 50-59 23 25 31 20 176 187 236 151 24 24 33 19 289 283 391 229 19 34 27 20 277 498 389 290 30 32 29 9 469 494 452 137 28 25 32 15 283 250 323 146 13 33 33 20 303 760 768 470
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 3 54 32 12 21 401 238 89 1 11 48 40 17 127 566 481 0 4 60 36 1 55 902 537 12 9 34 45 180 133 533 706 1 12 24 63 14 116 245 627 18 64 14 4 406 1,481 312 102
Marital/Residential Status
Unmarried and not cohabitating Married and/or cohabitating with partner 22 78 115 412 58 42 689 503 30 70 331 769 44 56 676 859 49 51 392 415 33 67 747 1,545
Number of Children
No children 1 to 2 3 to 5 6 or more 46 40 13 1 343 300 98 9 43 41 16 1 515 483 185 8 57 37 7 0 807 522 93 2 48 40 12 0 743 617 185 7 44 31 23 2 426 300 228 17 16 27 37 20 360 583 816 434
Table 2 - Socio-demographic Characteristics, Men
16
PART 1
PART 2
PART 3
socio-economic status), we also asked men about work-related stress, namely if they have ever have experienced stress or depression because of not having enough income or enough work. The hypothesis was that mens reports of experiencing such stress are more indicative of mens socioeconomic realities than only asking about monthly income or employment status. Results confirmed that work-related stress among men in the four middle-income and two low-income
countries in this study was common. A high percentage of men, regardless of their current employment status, report ever experiencing significant stress related to not having enough work or income. Between 34 percent and 88 percent of men in the survey sites report having experienced workrelated stress (Figure 1).14 Among currently unemployed men, between one-tenth and one-half of men interviewed report they considered leaving their family at least once because of work-related stress.
Brazil n=448
% N %
Chile n=426
N %
Croatia n=506
N %
India n=521
N %
Mexico n=383
N
Rwanda n=1311
% N
Age
18-24 25-34 35-49 50-59 22 23 33 23 98 102 147 101 15 22 37 26 62 93 159 112 19 32 30 19 95 162 150 98 26 36 30 7 137 190 160 38 13 21 45 20 51 79 173 78 17 38 30 15 219 497 395 200
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 50 11 19 20 222 49 87 90 7 10 34 50 29 42 144 211 4 0 57 39 21 0 287 196 23 6 38 33 119 30 201 175 16 16 34 34 59 62 128 131 53 30 9 9 689 393 112 117
Marital/Residential Status
Unmarried and not cohabitating Married and/or cohabitating with partner 44 56 198 250 20 80 62 248 47 53 233 265 20 80 103 422 11 89 41 339 27 73 355 956
Number of Children
No children 1 to 2 3 to 5 6 or more 2 70 25 2 7 227 82 6 3 75 21 1 10 235 65 2 1 80 19 0 2 199 46 1 4 66 28 2 15 273 115 10 0 48 45 8 0 159 149 25 10 31 40 19 124 401 517 248
Table 3 - Socio-demographic Characteristics, Women
17
Brazil 34 Chile 57 Croatia 53 India 62 Mexico 88

0 10 20 30 40 50 60 70 80 90 100
% of men reporting stress/depression related to work or income
Figure 1: Mens Reports of Work-related Stress
Discussion: Implications of Mens Work-Related Stress These initial findings suggest that given the prevailing, socially expected role of men to be providers and thus to have work, the indicators of economic stress used in IMAGES are more revealing about mens life experiences than only asking men about income and about their current employment status. Throughout this initial analysis of IMAGES results, this hypothesis was confirmed. On a number of key practices, men who reported experiencing economic or work-related stress are more likely to report depression, suicide ideation, previous arrests, and use of violence, among other practices.
seeking behaviors (Marsiglio, 1988; Kaufman, 1993; Rivers and Aggleton, 1998; Kimmel, 2000; Barker, 2000a; Barker and Ricardo, 2005). Sample survey research using standardized attitude scales, including the GEM Scale, (Pulerwitz and Barker, 2008) has found that adult and younger men who adhere to more rigid views about masculinity (e.g., believing that men need sex more than women, that men should dominate women, that women are responsible for domestic tasks, among others) are more likely to report use of violence against a partner, sexually transmitted infection, previous arrests and drug or alcohol use. To measure mens and womens gender-related attitudes, we applied the GEM Scale, a collection of attitude questions that has now been widely used in diverse settings and has consistently shown high rates of internal reliability.15 The scale also produced very high internal reliability in all the IMAGES study countries. The scale is not the same in all countries. Items have been added to the scale for cultural specificity and other items that show limited variation and limited contribution to the overall scale in that setting have not been used in the final data analysis.
Gender Attitudes
A key part of achieving gender equality is changing the social norms that men and women internalize and that influence their practices. Survey research with men and boys in numerous settings has shown how inequitable and rigid gender norms influence mens practices on a wide range of issues, including HIV/STI prevention, contraceptive use, use of physical violence (both against women and between men), domestic chores, caregiving, and health-
18
PART 1
PART 2
PART 3
Brazil GEM Scale Items by Domains Gender

Womans most important role is to take care of her home & cook Changing diapers, giving kids a bath & feeding kids are mothers responsibility A man should have the final word about decisions in his home 54 10 43 %
Chile
%
Croatia
%
India
%
Mexico
%
Rwanda
%
54 46 40
36 29 20
86 81
56 26 24
83 61 66
Violence
A woman should tolerate violence in order to keep her family together There are times when a woman deserves to be beaten 4 6 12 68 65 54 21
Sexuality
Men need sex more than women do Men dont talk about sex, they just do it Men are always ready to have sex I would never have a gay friend Its important for men to have male friends to talk about his problems 50 49 54 46 90 32 25 34 31 57 58 61 27 31 42 29 70 58 54
Masculinities
To be a man, you need to be tough Men should be embarrassed if unable to get an erection If someone insults me, I will defend my reputation, with force if I have to 44 37 38 46 69 62 86 91 92 8 13 38 19 59 35
Reproductive Health
It is a womans responsibility to avoid getting pregnant I would be outraged if my wife asked me to use a condom Either a man or a woman can suggest using a condom If a guy gets women pregnant, child is responsibility of both Man/woman should know what his/her partner likes during sex The participation of the father is important in raising children Couple should decide together if they want to have children A man and a woman should decide together what type of contraceptive to use Total # of GEM items included in scale 36 21 11 47 33 90 98 97 98 98 15 16 12 92 13 40 47 12 22 11 49 39 13
indicates that these items were not used in the final construction of the GEMS in this country
Table 4 - Percent of Men Who Agree or Partially Agree with GEM Scale Items by Country
19
In this section, we present the GEM Scale results focusing on responses to individual questions. In subsequent sections, the GEM Scale is examined as a variable as it relates to mens reports of their practices in terms of caregiving, sexual experiences, use of violence, and health-seeking behaviors, among others.16 In Annex III, we present the GEM scores for each country trichotomized into categories of men holding low, medium and high equity genderrelated attitudes. Men showed tremendous variation in their genderrelated attitudes, with India and Rwanda showing the most inequitable attitudes. Table 4 presents the responses to each attitude question by country. In terms of roles in the home, sexuality, reproductive health and gender-based violence, Rwandan and Indian men consistently supported the least equitable norms among the settings studied. For example, for the statement changing diapers, giving kids a bath and feeding kids are the mothers responsibility, only 10 percent of men agreed in Brazil whereas 61 percent in Rwanda and more than 80 percent in India agreed with the statement. In India and Rwanda, men also showed high acceptability of mens use of violence against women, and in both countries, a majority
affirmed the belief that men need sex more than women do. Men with higher educational attainment and married men had more equitable attitudes; unmarried men had the least equitable attitudes. In terms of factors associated with whether men were more or less equitable, unmarried men (with or without a partner) have the least equitable gender attitudes across all the countries. In terms of age, we see mixed trends. In some countries, younger men show more equitable views. In other countries, men over the age of 50 are more equitable than their younger counterparts. Homophobic attitudes were common although varied tremendously by context. Questions about mens attitudes related to homophobia or acceptance of sexual diversity were added to the one GEM Scale question related to homophobia (I would never have a gay friend ). Men who said they would be ashamed to have a gay son ranged from a low of 43 percent of men in Brazil to a high of 92 percent in India. A slightly lower, but still high proportion of men said that being around homosexual men makes them uncomfortable, ranging from a low of 21 percent of men in Brazil to a
Brazil 21 43 Chile 55 44 Croatia 63 63 India
Being around homosexual men makes me uncomfortable I would be ashamed if I had a homosexual son
89 92
0 10 20 30 40 50 60 70 80 90 100
% of men who agree
Figure 2: Mens Attitudes about Homosexuality
20
PART 1
high of 89 percent in India. Among the four countries where these questions were asked, Brazil and Chile had the least homophobic responses, while Croatia and India had more homophobic responses. Although not presented in detail here, analysis found that younger men were less likely to hold homophobic attitudes in the cases of Brazil, Chile and Croatia (at statistically significant levels); in Brazil, Croatia and India, men with higher levels of education were less likely to hold homophobic attitudes (at statistically significant levels).
PART 2
PART 3
Discussion: Gender Attitudes The fact that unmarried men had the least equitable attitudes suggests that men who are married may learn to soften or modify their attitudes as they acquire experience in co-habitating relationships with women. In addition, the finding that in some countries, older men had more gender-equitable attitudes than men in the middle-age groupings further affirms the importance of men gaining experience in co-habitating relationships and its influence on mens attitudes. Years of living in partner relationships and the daily negotiation required may, from a developmental perspective, lead some men to become more gender equitable in their attitudes. In all the countries, men with higher educational attainment (completed primary school and at least some secondary education) had more equitable attitudes than those with less education. This suggests that secondary schools may be a space where more rights education takes place.
Richard Lewisohn/Influential Men
21
Brazil
Men
Women
86 88 Chile 88 90 Croatia 93 91 India 67 70 Mexico 89 88

0 10 20 30 40 50 60 70 80 90 100
% who agree
Figure 3: Mens and Womens Reports about Speaking with Partner about a Personal Problem in Previous Month
Mens Practices and Lives

Relationship Dynamics and Domestic Duties Globally, an unequal work-life divide endures where men are generally expected to be providers and breadwinners while women and girls are generally expected to provide care for children and other dependents and be chiefly responsible for reproductive aspects of family life. A recent multi-country study including lower, middle and higher income countries found that the mean time spent on unpaid care work by women ranges from two to 10 times that of men (Budlender, 2008). These realities persist even as women have begun working outside the home in larger and unprecedented numbers and as their roles have changed in households and political life, albeit with tremendous variation by region and social class.
coincide? Are respondents who are co-habitating heterosexual couples satisfied with the current division of care work, including care of children? In terms of relationship dynamics, key IMAGES questions focused on division of household duties, spousal communication, overall relationship satisfaction and sexual satisfaction
Both men and women report frequent communication about personal problems. The vast majority of men in co-habitating heterosexual relationships say they have spoken with their female partner about a personal problem they were facing in the last month, ranging from 66 percent in India to 93 percent in Croatia; women reported nearly identical rates. For most of the countries, couple communication varied little by age or workrelated stress, but did vary in terms of educational What do men and women in the six countries stud- attainment and gender norms. In Croatia and Mexico, men with higher levels of education were ied say about these issues? Are men doing their share of care work? And how do mens and womens significantly more likely to report talking to partners in the past month about personal problems. In perceptions of who does the care work differ or
22
PART 1
India, men facing work-related stress were more likely to have talked with a partner about personal problems. In nearly all of the countries, men who have more gender-equitable attitudes were more likely to report having spoken with a partner about personal problems in the last month. According to mens own reports of their practices, younger men, men with more education and men who saw their fathers do domestic work are more likely to carry out domestic duties. Nearly half of men in all sites say that they play an equal or greater role in one or more household duties with the notable exception of India, where only 16 percent of men reported that they played an equal or greater role in household duties. These household or domestic duties included washing clothes, repairing the house, buying food, cleaning the house, cleaning the bathroom or toilet, preparing food and paying the bills.17 Not surprisingly, the tasks that men said they play an equal or greater role in are those tradi-
PART 2
PART 3
tionally associated with men namely repairing the house, paying bills and buying groceries. Womens reports that their male partner played a greater or equal role in at least one household task were consistently lower than mens own reports, with the exception of India where women and men agreed that men seldom play an equal or greater role in any household duties (Figure 4). In terms of factors associated with mens involvement in domestic activities, there was a statistically significant association with age in Chile, Croatia and Rwanda. Younger men were more likely to say they carried out an equal or greater role in at least one domestic activity (and the same trend was seen in India and Mexico, but not at a statistically significant level). Men with higher educational levels were more likely to report participation in domestic activities in Brazil and Chile, where the difference was significant, although the same pattern was seen in
Brazil 26 Chile 52 35 Croatia
Men
Women
60
61 47 India 16 18 Mexico 54 47 Rwanda 52 23

0 10 20 30 40 50 60 70
% reporting that man plays equal or greater role in one or more domestic duties
Figure 4: Mens and Womens Reports of Mens Participation in Domestic Duties
23
Repair House
Brazil n=424 Chile n=655 Croatia n=791 India n=810 Mexico n=547 Rwanda n=1949
Pay Bills
Buy Groceries 69 65 74 39 56 63
Clean House
Prepare Food
Clean Bathroom 49 28 27 6 36 32
Wash Clothes
88 89 96 66 92 95
76 74 76 89 81 89
49 36 43 7 42 33
45 37 39 5 31 33
38 26 17 8 27 41
Table 5: Mens Reports of Equal or Greater Participation in Domestic Duties, in Order from Most to Least Common Overall
Brazil 58 60 Chile* 43 59 Croatia* 52
Men whose fathers DID NOT participate in domestic duties Men whose fathers DID participate in domestic duties
66 India 13 18 Mexico* 44 69 Rwanda* 46 60

0 10 20 30 40 50 60 70 80
*statistically significant at the p < .05 level
% who play an equal or greater role in domestic duties
Figure 5: Links between Fathers and Sons Participation in Domestic Duties (Defined as Playing an Equal or Greater Role in One or More Duties) nearly every country (not at statistically significant levels). Men who showed more gender-equitable attitudes were more likely in all the countries except India to participate in domestic activities. Additionally, in all the sites, men who reported that their fathers participated in domestic activities were more likely to report participation in domestic activities (Figure 5).
24
PART 1
Men report higher levels of sexual and relationship satisfaction than women. Women who say their partners do domestic work are more sexually satisfied. Men reported relatively high rates of sexual satisfaction with their current stable partners ranging from 77 percent in Croatia to 98 percent in India. Women reported significantly lower rates of sexual satisfaction than men in Brazil, Chile, Mexico and Rwanda, while reporting roughly equal rates of sexual satisfaction as men in the case of Croatia and India. Younger men reported higher levels of sexual satisfaction than did older men in Chile, Croatia and Mexico. Men with higher educational levels reported higher levels of sexual satisfaction in Brazil, Mexico and Rwanda. Work-related stress was not associated with sexual satisfaction except in the case of Croatia, where men who reported facing stress related to work and income reported lower levels of sexual satisfaction with their current partner. In all countries except India, men who reported more gender-equitable attitudes were more likely to
PART 2
PART 3
report being sexually satisfied with their current female partner. Furthermore, in Brazil, Croatia and India, women who reported that their male partner carried out an equal or greater role in at least one domestic activity reported higher levels of overall relationship satisfaction and higher levels of sexual satisfaction (Figure 8). Men who reported higher levels of communication with their partners about personal problems in Chile, Croatia and Mexico were also not surprisingly more likely to report that they were more sexually satisfied with their current female partners (Figure 7). Not surprisingly, across all the countries included, the vast majority of men say they are satisfied with the current and highly unequal division of household duties. Mens reports of satisfaction ranged from 90 percent in Brazil to 98 percent in India. However, womens reports of satisfaction with the current division of domestic duties are consistently lower, but only slightly (Figure 9).
Brazil 77 Chile
Men
Women
92
86 75 Croatia 77 78 India 98 96 Mexico 95 86 Rwanda 86 60

0 10 20 30 40 50 60 70 80 90 100
% reporting sexual satisfaction with current partner
Figure 6: Mens and Womens Reports of Sexual Satisfaction with their Current Partner
25
Discussion: Relationship Dynamics and Domestic Duties The overall finding in terms of relationship dynamics is that men with more education, younger men and men with more equitable views are more likely to participate (or say they participate) in household work and to report higher levels of sexual and overall relationship satisfaction. The findings also confirm, for some sites, a clear association for women between their male partners participation in household work and their sexual satisfaction and overall satisfaction with their relationship. These results affirm that communication and greater sharing of household tasks are associated with greater overall relationship satisfaction and sexual satisfaction. We should be careful in making any simplistic association between equality in domestic tasks and sexual or overall relationship satisfaction (not to mention the challenges of measuring sexual satisfaction across cultural contexts), but these results do suggest that cooperative relationships are generally cooperative across several areas of intimate and household life and that women and men appreciate this cooperation in all its forms.
These results also suggest that women on aggregate have generally conformed to mens limited participation in domestic activities but are also appreciative when their male partners participate more in such activities. Indeed, given the obvious inequalities in the division of household tasks involving cleaning, food preparation and washing, womens reported dissatisfaction is less than we might have imagined. This likely means that women have accepted or internalized these inequalities stemming from the persistent gendered division of labor, do not believe these arrangements will change anytime soon or do not expect any more from their male partners in terms of domestic tasks. In addition, these results show the importance of mens participation in domestic work as an influential factor in changing behaviors from one generation to another. Men who grow up in homes where men regularly participate and assume responsibility for care work, including the care of children, are more likely to see such practices as mens work and in turn to carry out such practices when they form their own families.
Brazil
Has not talked to partner about problems in last month
Has talked to partner about problems in last month
92 93 Chile* 67 89 Croatia* 47 80 India 98 98 Mexico* 82 96

0 10 20 30 40 50 60 70 80 90 100
% who report sexual satisfaction
Figure 7: Mens Reports of Sexual Satisfaction and their Reports of Couple Communication
26
PART 1
PART 2
PART 3
Brazil* 43
Has not talked to partner about problems in last month
Has talked to partner about problems in last month
82 Chile* 55 78 Croatia* 48 84 India 95 96 Mexico* 63 90

0 10 20 30 40 50 60 70 80 90 100
% who report sexual satisfaction
Figure 8: Womens Reports of Sexual Satisfaction and their Reports of Couple Communication
Brazil
Men
Women
90 88 Chile 94 80 Croatia 93 80 India 98 97 Mexico 94 88

0 10 20 30 40 50 60 70 80 90 100
% reporting satisfaction with current division of duties
Figure 9: Mens and Womens Reports of Satisfaction with Current Division of Household Duties
27
Parenting and Involvement in Childbirth What are mens and womens reports of mens involvement in the care of children? Key IMAGES questions focused on mens participation in prenatal visits, presence during childbirth, involvement in the daily care of children and use of paternity leave.
birth, with some notable exceptions. In Chile, nearly 50 percent of men reported that they were present in the delivery room during the birth of their last child, followed by 24 percent in Mexico and 17 percent in Croatia. These results also must be contextualized by setting. In some settings, official public health policy supports womens rights to In most settings, the majority of men were neither in the have a person of her choice present during delivery room nor the hospital for the birth of their last childbirth (Brazil, Chile and Croatia), while in child, although younger men and men with more eduMexico public hospitals do not permit men (fathers) cation are more likely to be present during childbirth.18 to be present during birth even if the father and In many middle-income countries, particularly in mother desire. Brazil has a national policy supportWestern Europe, North America and Australia, ing a womans right to have someone of their choice national health systems have made efforts to involve (presumably the father) present during childbirth, men to a greater extent in maternal health and duryet implementation has been far less than universal ing childbirth. The most common engagement is to as more than half of men reported they were not invite or encourage men to be present during prena- present for the birth. Chile shows a dramatic tal checkups and be present in the delivery room.19,20 generational shift in terms of mens presence in the delivery room during childbirth (Figure 11), largely In spite of increasing efforts to promote mens due to a national policy that aims to humanize the involvement during childbirth, the majority of men birth process, which encourages women to have a across the countries surveyed are neither present in male partner or other person of their choice present the delivery room nor in the hospital during childduring birth at public maternity wards.
Brazil (n=407) 7 Chile (n=534) 50 Croatia (n=531) 17 India (n=762) 2 Mexico (n=464) 24 Rwanda (n=1801) 11
0 10 20
In delivery room
Elsewhere in hospital
Not present
39
54
22
28
28
54
71
28
73
40
30 40 50 60 70
49
80 90 100
% of men at specified location for birth of last child
Figure 10: Mens Presence During Birth of Last Child
28
PART 1
PART 2
PART 3
18-24 90 25-34 62 35-49 52 50-59 31

0 10 20 30 40 50 60 70 80 90 100
% present in the delivery room for birth of last child
Figure 11: Mens Presence in the Delivery Room by Age in Chile (statistically significant at the p < .001 level)
In terms of factors associated with being present (either in the hospital or the delivery room), younger men (mostly in the 25-34 year old range) were more likely to be present in the cases of Chile, Croatia and India, with similar (but not statistically significant) patterns seen in Mexico and Rwanda. More educated men were more likely to report being present during childbirth in all the countries (at statistically significant levels in all but India). The majority of men and women say that men were present during at least one prenatal visit. Younger men and more educated men were more likely to be present during prenatal visits. While men are not present during childbirth, they are accompanying women to prenatal visits at high rates in all the countries, ranging from 78 percent in Brazil to 92 percent in Mexico (referring to the percent of men who reported that they accompanied their female partner for at least one prenatal visit during the most recent pregnancy). Womens reports of mens participation in prenatal visits are lower in all of the countries, although more than half of women report that men were involved in prenatal visits in their most recent pregnancy (Figure 12). Men with higher educational levels were more likely to accompany a partner to prenatal visits in Brazil,
Chile, Croatia and Mexico, as were men who showed more equitable views about gender. Younger men were more likely to accompany a partner to prenatal visits in the case of Chile (although similar but statistically insignificant differences were seen in all the other countries). Men are taking few days of paid or unpaid paternity leave, although younger men and men with more education are more likely to take leave. Expanding paternity leave particularly paid paternity leave has been promoted as a way to increase
David Snyder/ICRW
29
Brazil 78 59 Chile
Men
Women
86 72 Croatia 92 75 India 91 86 Mexico 92 80

0 10 20 30 40 50 60 70 80 90 100
% who say yes, man accompanied
Figure 12: Mens and Womens Reports of Mens Accompaniment During Prenatal Visits During Most Recent Pregnancy
mens participation in the care of children, and included in recommendations from the 48th Session of the Commission on the Status of Women (CSW) in 2004. Most countries with the notable exceptions of some Western European countries, Canada, Australia and New Zealand continue to offer only a few days or a week at most of paid or unpaid paternity leave. This limited paternity leave continues to reinforce the notion that men are helpers when it comes to the care of children, both perpetuating unequal care work patterns and discouraging mens involvement in their childrens lives. These trends persist in spite of widely reported positive experiences in terms of mens use of leave and womens and mens reported satisfaction as men take on increased involvement in the care of children. Among the countries included in the study, Brazil and Chile have laws that provide men with five days paid leave (paid through national social security taxes), Mexico offers none (although Mexico City, not included in the survey, has a municipal provision guaranteeing men 10 days paid paternity
leave), India has none (although national government employees have 15 days paid paternity leave) and Rwanda has none. Croatia offers paid paternity leave of up to three months and provides for some flexibility of transferring leave between mothers and fathers, although the results here affirm that about 37 percent of men take this leave but for far less than the allowed time. Among men who took leave, the duration ranged from an average of 3.36 to 11.49 days of paid leave and from an average of 3.8 to 10 days unpaid leave. Younger men are more likely to take leave in the case of Chile and Croatia (although similar but statistically insignificant differences were seen in India, Mexico and Rwanda). Men with higher educational levels are more likely to take leave in the case of India and Rwanda (with similar but statistically insignificant differences seen in Brazil and Chile). Men who reported work-related stress were less likely to take leave in Brazil and India (with similar but statistically insignificant trends seen in Chile and Croatia). Men with more gender-equitable atti-
30
PART 1
tudes are more likely to take leave in the case of Croatia and Rwanda (with trends seen in the same direction in Brazil, Chile and Mexico but not at statistically significant levels). Close to half of men with children say they are involved in some daily caregiving. Unemployed men are dramatically more likely to participate in the care of children than employed men. Mens reports of providing daily care for children ranged from 36 percent in Chile to 63 percent in Croatia.21 However, only 10 percent to 31 percent of women reported that their male partners provide daily care for children (Figure 14). While the rate of mens selfreported participation in the daily care of children in Croatia seems high, it is important to note that very few men there reported caring for children by themselves. That is, in Croatia as in most settings, most men who report participating in the daily care of children report equally or done together with my partner. Women report that mens most common participation in child care (for those who have children age
PART 2
PART 3
four and younger) is playing with children, which was nearly 50 percent and above in all the countries (Figure 16). Women reported that men are much less likely to be involved in cooking and changing diapers, with the exception of Chile, where more than 90 percent of women (with children younger than age four) say that men both play with and prepare food for young children several times a week or on a daily basis. In terms of factors associated with mens daily participation in the care of children, the most dramatic association was mens employment status. In all the countries, men who are currently unemployed reported significantly more involvement in the daily care of children when compared to currently employed men up to 50 percent higher in some countries (Figure 17). This suggests that men with more time (albeit probably involuntarily) are assuming a greater share of the care of children and more than is commonly recognized.22 There were few other clear associations in terms of mens participation in the daily care of children by age, education levels or work-related stress that were
Brazil (n=342) 38 Chile (n=470) 77 Croatia (n=462) 60 India (n=589) 31 Mexico (n=403) 34 Rwanda (n=1227) 52
0 10 20 30 40 50 60
No Leave
Unpaid Leave
Paid Leave
61
21
37
44
25
13
53
37
70 80 90
11
100
% who report type of leave
Figure 13: Parental Leave After Birth of Most Recent Child Among Currently Working Men
31
Brazil 10 Chile 36 10 Croatia
Mens Reports
Womens Reports (of Men)
39
63 17 India 37 18 Mexico 46 31
0 10 20 30 40 50 60
% who participate in the daily care of a child
Figure 14: Mens and Womens Reports of Mens Participation in the Daily Care of a Child
Brazil 83 46 52 Chile 70 27 39 Croatia 76 45 52 India 46 18 19 Mexico 64 28 35

0 10 20 30 40 50 60 70 80 90 100 Change diapers for child age 0-4 Cook for child age 0-4 Play with child age 0-4
% who report doing activity several times a week or more
Figure 15: Mens Reports of Care of Children Age 4 and Younger
32
PART 1
consistent across sites. In Brazil, men with higher levels of education were more likely to participate in the daily care of children (with slight, but statistically insignificant differences seen in the same direction in all countries). In Croatia, men who reported work-related stress were more likely to participate in the daily care of children. Men with more gender-equitable attitudes were more likely to report being involved in the daily care of children in Chile and Croatia, with statistically insignificant differences in the same direction seen in Brazil.
PART 2
PART 3
Discussion: Mens Involvement in Parenting and Childbirth As in the case of mens participation in domestic activities, younger men, men with more education and men with more equitable attitudes are more likely to report participating both in childbirth and in the daily care of children. Among the different forms of participation in the lives of their children, mens participation in prenatal visits is the most frequent of these. Indeed, if mens participation during childbirth and in the daily care of children is lagging, mens participation in prenatal visits seems to be an increasingly widespread and accepted practice and may provide a key point of entry for engaging men in caregiving, child health and other aspects of maternal health.
Brazil 72 25 42 Chile 95 91 Croatia 71 42 51 India 74 30 29 Mexico 46 15 23

0 10 20 30 40 50 60 70 80 90 100 Change diapers for child age 0-4 Cook for child age 0-4 Play with child age 0-4
% who report that partner does activity several times a week or more
Figure 16: Womens Reports of Mens Care of Children Age 4 and Younger
33
Brazil 53 35 Chile* 57 34 Croatia
Not Working
Working
78
61 India 42 37 Mexico*
68
44
0 10 20 30 40 50 60 70 80
% who participate in the daily care of a child
Figure 17: Mens Employment Status and Daily Care of a Child
Looking specifically at paternity leave, a relatively high percentage of men are taking leave particularly taking into account that paternity leave is still a relatively new phenomenon in many settings and is not paid, nor guaranteed as a labor right in the case of India, Mexico and Rwanda. However, the amount of leave that men are taking in the cities included in IMAGES is still too limited to make a major difference in terms of changes in the inequality in care work.
34
Nicholas Pitt/Influential Men
It is also important to note the discrepancies between womens reports of mens child care practices and mens own reports. Based on these discrepancies, we conclude either that men are over-reporting their participation or that women are under-reporting mens participation or not recognizing this participation (or more likely, both). In some cases, though, womens reports of mens participation in the care of children were higher than mens own reports. All of these issues point to the complexity of measuring mens participation in child care, as other researchers have (NCOFF, 2002; Brown and Chevannes, 1998). In addition, IMAGES did not include questions on other ways that men may be involved with their children providing guidance or financial support from afar if they have migrated for work, accompanying children to school or on other outings or activities outside the home, or the extra hours they may work to pay for school fees or other costs associated with children. These are all ways in which men may provide care for children that are not frequently counted as care work, either by female partners, by men themselves or by researchers.
PART 1
Health Practices and Vulnerabilities Research consistently affirms that practices related to health and well-being, living conditions and access to health services are closely linked to societal constructs of masculinity and femininity and the power and resource inequalities between men and women and within specific groups of men and women. Womens health-related vulnerabilities have been highlighted in key events and documents (WHO, 2009).
PART 2
PART 3
particularly economically and socially marginalized men such as men who migrate for work and men in hazardous occupations face specific, gendered health risks (see WHO, 2007).
These issues show up in global health statistics. For example, according to the World Banks 2005 World Development Indicators, women have higher life expectancies than men in all but four countries (World Bank, 2005). Substance abuse and addiction In recent years, there also has been a growing aware- are predominantly male phenomena worldwide. ness of the ways in which gender norms interact Recent evidence shows that death and disability with income inequalities to create specific gendered rates related to alcohol and substance abuse are vulnerabilities for men in terms of health and wellconsiderably higher for men than for women (WHO, being (Bannon and Correia, 2006; Courtenay, 2000). 2004; Pyne, et al., 2002). These gendered health Research has affirmed, for example, that culturally patterns have high costs for women, children, socidominant forms of masculinity, which often urge eties and men themselves. Mens shorter lives mean men to practice strict emotional control and cultiless productivity and curtailed participation as vate a sense of invulnerability, serve as barriers to caregivers and intimate partners. Mens alcohol and health- and help-seeking behavior or encourage substance abuse, as seen in the results presented some men to engage in practices detrimental to their here, is a factor in mens use of intimate partner own health and that of their families. Other research violence, and in some studies has been found to be has affirmed the ways in specific groups of men associated with violence against children. To further
Brazil 69 20 Chile 39 17 Croatia 42 24 India 23 Mexico 34 2

0 10 20 30 40 50 60 70 80 Mens regular abuse of alcohol Womens regular abuse of alcohol
% who report their own regular abuse of alcohol
Figure 18: Mens Reports of Regular Abuse of Alcohol Compared to Womens Reports of (Their Own) Regular Abuse of Alcohol (five or more drinks on one occasion once monthly or more)
35
examine these issues, the IMAGES questionnaire included items related to mens physical, mental, and sexual and reproductive health.
had prostate exams ranges from 20 percent in Mexico to 42 percent in Croatia, a strikingly low rate given the specific risk prostate cancer poses to men over age 40 and the existence of campaigns in many countries enjoining men to seek regular exams Mens alcohol abuse is much higher than that of women; men with more inequitable gender norms are (Figure 19). Even men in the 50-59 age group did not surpass a 50 percent likelihood of having had a more likely to abuse alcohol. Mens rates of regular abuse of alcohol defined as prostate exam in any of the countries. having five or more drinks in one night on a once Men had relatively low rates of HIV testing and conmonthly or greater basis vary from 23 percent in sistently lower than womens rates of testing (with the India to 69 percent in Brazil and are significantly exception of Rwanda, where mens seeking of HIV higher than womens reported alcohol abuse (see testing was high). Higher education and support for Figure 18). more equitable gender norms are associated with Younger men are more regular alcohol abusers than mens greater likelihood to have been tested for HIV. Mens rate of testing in Rwanda was 87 percent, as older men in Chile, Croatia, and Mexico. In Chile, Croatia, and Mexico, men holding more inequitable compared to a low of 11 percent of men in Croatia (Figure 20). Education level was a statistically sigviews about gender were by far the most regular abusers of alcohol. In Chile, for example, 69 percent nificant influence on the likelihood of men having been tested in four countries (Chile, Croatia, Mexico of the men holding the least equitable gender attiand Rwanda). Men with more gender-equitable attitudes regularly abuse alcohol compared to 35 tudes were also more likely to have been tested in percent for men who hold the most equitable attiCroatia, India and Mexico. Womens rates of HIV tudes. The same figure in Mexico is 50 percent as testing were higher than men everywhere but Croacompared to a 34 percent average among the three tia. These higher rates for women are likely due to cities included. In India, men with incomplete primary education and those who reported work stress routine testing of women as part of prenatal care. were more likely to report alcohol abuse. Both women and men report that men are frequently Men seem reluctant to seek regular prostate exams, in involved in the decision to terminate a pregnancy when women inform men of the pregnancy. spite of campaigns to encourage men age 40 and Between 11 percent to 27 percent of women interolder to do so. viewed say they have ever sought an abortion. Mens The percentage of men age 40 and older who have
Brazil 33 Croatia 42 India 22 Mexico 20

0 10 20 30 40
% of men older than age 40 who have ever sought a prostate exam
Figure 19: Men Older than age 40 who Have Ever Sought a Prostate Exam
36
PART 1
PART 2
PART 3
Brazil 35 65 Chile 29 49 Croatia 11 9 India 13 37 Mexico 27 30 Rwanda
Men
Women
87 89
0 10 20 30 40 50 60 70 80 90 100
% who report ever having been tested for HIV
Figure 20: Mens and Womens Reports of Having Ever Sought an HIV Test
reports of having a female partner who ever sought an abortion are generally much lower, however (except in Brazil and Croatia). Among those men who reported that a female partner had ever sought an abortion, high proportions (30 percent to 71 percent) say that they were involved in the decision to terminate the pregnancy. Similarly, high proportions of women affirmed this (39 percent to 92 percent). This suggests that if the man is informed about the pregnancy, which is clearly not always the case, both women and men affirm in relatively similar percentages that men participate to a fairly substantial degree in the decision to terminate a pregnancy. While men tend to report high self-esteem, men also show relatively high levels of negative feelings and depression, with exceptions in Croatia and India. Men in Brazil, Chile and Mexico affirmed three items measuring positive self-esteem at levels consistently
Richard Lewisohn/Influential Men
37
Brazil 18 23 Chile 6 11 Croatia 20
Men
Women
27 India 2 21 Mexico 12 25
0 10 20 30
% who report that they (women) or partner (men) ever terminated a pregnancy
Figure 21: Reports of Induced Abortion
Brazil 30 39 Chile 65 47 Croatia 71 41 India 59
Men
Women
92 Mexico 18 10
0 10 20 30 40 50 60 70 80 90 100
% who report that man was involved in decision to terminate pregnancy (of those who have had - women - or whose partners have had - men - an abortion)
Figure 22: Mens and Womens Reports of Mens Involvement in Abortion Decision-Making (of Those Who Report That They or a Partner Ever Sought an Abortion)
38
PART 1
PART 2
PART 3
exceeding 90 percent (Figure 23). Significant minori- experiences of depression are related strongly with ties of their Indian and Croatian peers, however, did work- and income-related stress. In four of the five not affirm these positive statements. countries where questions about depression were included, men who report work-related stress are When posed with the complementary negative state- more likely to report depression (with the associaments of self-esteem (Figure 24), the trends were tion being statistically significant). similar. In no country did more than 10 percent of respondents agree that I feel that my life is of no A substantial proportion of men in all countries report use to anyone, but significant percentages in all recent suicidal thoughts. The link with suicidal countries but Mexico agreed that I feel inferior thoughts and work- or income-related stress is strong sometimes when I am together with friends. in four countries. In 2004, the WHO World Health Report showed that of When asked about feeling depressed, however, 93 countries reporting gender disaggregated suicide men show signs of more vulnerability. The rates of rates, all but two had higher male suicide rates than experiencing depression at least once in the last female (WHO 2004). IMAGES data also affirm the month ranged from 9 percent in Brazil to a high of high frequency at which men hold suicidal thoughts. 33 percent in Croatia.23 The data further show that Among the study countries, the percentages of respon-
Im happy with my body
Brazil
I have a lot to be proud of on the whole
I have a good sex life
88 94 94 Chile 95 95 83 Croatia 72 74 65 India 95 69 71 Mexico 96 97 92

0 10 20 30 40 50 60 70 80 90 100
% who agree
Figure 23: Mens Affirmations of Positive Self-Esteem
39
dents who reported having suicidal thoughts someof stress about work or income are strongly associtimes or often in the last month ranged from 1 percent ated with the likelihood of having suicidal thoughts. in Mexico to 5 percent in Croatia (Figure 26). Again, in three of the of the five study countries where the question was asked, those men who report As with depression, male respondents experiences feelings of stress about work or income are more
I feel my life is of no use to anyone
I feel inferior sometimes when I am together with friends
16 India 7 29 Mexico 6 6
0 10 20 30
% who agree
Figure 24: Mens Negative Statements of Well-Being

0 10 20 30
% of men reporting feeling depressed sometimes or often in last month
Figure 25: Mens Reports of Feeling Depressed (Sometimes or Often in the Last Month)
40
PART 1
PART 2
PART 3

0 1 2 3 4 5
% of men who report having suicidal thoughts sometimes or often in the last month
Figure 26: Mens Reports of Suicidal Thoughts (Sometimes or Often Thinking About Suicide in the Past Month)
Brazil* 1 3 Chile 3
No, I dont frequently feel stressed or depressed about not having enough income
Yes, I frequently feel stressed or depressed about not having enough income
4 Croatia* 2 7 India* 1 4 Mexico 0 1

0 1 2 3 4 5 6 7 8
% who experienced suicidal thoughts sometimes or often in the last month
Figure 27: Association Between Mens Reports of Work-Related Stress and Suicidal Thoughts
likely to report harboring suicidal thoughts at a statistically significant level (Figure 27). Men seek help at far lower rates than women, despite mens high levels of depression and suicidal thoughts. In no country did more than two-thirds of men report seeking help in cases of sadness, disappoint-
ment or frustration (Figure 28). In the two countries where womens data are available on the same question, the gendered differences are stark. For example, only 11 percent of Indian men seek help or support in these situations as compared to 93 percent of their female counterparts.
41
Brazil 52 Chile 55 Croatia 67
Men
Women (only available in two countries)
83 India 11 93 Mexico 48
0 10 20 30 40 50 60 70 80 90 100
% who report seeking help
Figure 28: Mens and Womens Reports of Seeking Help When They Feel Sad, Disappointed or Frustrated Discussion: Health Practices and Vulnerabilities The results in this section affirm a strong association between holding inequitable views about gender (and rigid norms about what it means to be man) and mens negative health practices and vulnerabilities. These results also affirm that a key factor associated with mens health vulnerabilities is their work-related stress. These findings echo the evidence in the literature that conforming to stoic and rigid notions of masculinity for example, that a man must provide financially for his family and should not show vulnerabilities or seek help contribute to suicidal behavior and depression (Moller-Leimkuhler, 2003; Emslie, et al. 2006).
42
David Isaksson/Influential Men
PART 1
Violence and Criminal Practices The IMAGES questions on violence were informed by existing literature affirming the linkages between mens use of violence, their experiences of violence during childhood, and socially prevalent norms related to masculinities (Connell, 2000). Violence is one of the most extreme manifestations of power inequalities and is both experienced and used by many men to exert control and dominance over women, children and other men. In formulating
PART 2
PART 3
questions on violence, we took into account this previous research showing these linkages between mens experiences of violence and their use of violence, in particularly childhood experiences of violence in multiple forms, as well as attitudes about gender and manhood. In recent years, the amount of research on violence against women has increased tremendously, accompanied by the development of more sophisticated
Reports ever perpetrating physical violence against partner
Reports ever perpetrating physical violence against partner in the last year
% Brazil Chile Croatia India Mexico Rwanda 24 30 33 37 17 39
n 738 1147 1369 929 989 1709
% 11
n 738
19
929
1709
Table 6: Mens Reports of Physical Violence Against a Female Partner
Survivor of physical violence ever % n Brazil Chile Croatia India Mexico Rwanda 28 31 38 31 31 41 445 408 465 524 378 1086
Survivor of physical violence in the last year % n 8 445
19
525
1090
Table 7: Womens Reports of Physical Violence by an Intimate Male Partner
43
research instruments and more appropriate ethical guidelines, which IMAGES followed. Most notably, the WHO multi-country study on violence against women, and other surveys based on that, have provided representative, household data on womens reports of numerous forms of violence from male partners and the factors associated with that violence (Garca-Moreno, et al., 2005). Although there also have been a number of studies on mens experiences of and use of violence (against other men and against female partners), few studies, if any, are on par with the scale of the WHO multicountry study of women, and fewer still in the context of low- and middle-income countries.
on intimate partner violence in heterosexual couples, including the association of violence experienced during childhood, perpetration of sexual violence against women (either intimate partners or strangers), mens reports of their involvement in criminal activities and use of violence against other men.
Men reported lifetime rates of using intimate partner violence ranging from 25 percent to 40 percent with women reporting slightly higher rates of victimization. The WHO multi-country study (Garca-Moreno, et al., 2005) found that womens lifetime reports of physical intimate partner violence (IPV) were between 10 percent and 70 percent among the 10 countries studied.24 In IMAGES, IPV was measured Accordingly, IMAGES provides population-based data using a slightly modified version of the WHO methodology. Men were asked about particular on the prevalence of different acts of violence and types of physically violent acts perpetrated against explores more in-depth the factors associated with their female partners, and women were asked about violence against women from both mens and womens perspectives. Key IMAGES questions focused their experiences of the same forms of violence.25,26
Men
Women
38 India 37 31 Mexico 17 31 Rwanda 39 41

0 10 20 30 40
% who report perpetrating (men) or experiencing (women) violence
Figure 29: Mens Reports of Using IPV Against a Female Partner and Womens Reports of Experiencing IPV from a Male Partner (Lifetime)
44
PART 1
Mens lifetime reports of using IPV varied from around 25 percent to almost 40 percent, with the exception of Mexico where fewer than 20 percent of men reported ever use of IPV (Table 6).27 Brazil and Chile show rates below 30 percent while India and Rwanda have the highest rates at around 38 percent. Reports of IPV use in the last year range from 2 percent in Rwanda to almost 19 percent in India and 11 percent in Brazil. The 2 percent rate reported by men in Rwanda is far lower than the rate of 9 percent of women who reported experiencing IPV in the last year, suggesting that men may be under-reporting IPV in Rwanda. This may be due to changing social acceptability of violence as exhibited in a new law to prevent gender-based violence and increased efforts by NGOs and the government on the issue. Womens lifetime reports of IPV in IMAGES sites varied from 27 percent to 41 percent, only slightly higher than mens reports everywhere but India, where women reported lower rates than men, and Mexico, where men reported significantly lower rates than women (Table 7 and Figure 29). Overall, as shown in Annex V, the lifetime levels of IPV found in IMAGES are similar to those found in previous large population-based surveys in the six countries. It is important to mention that in four countries (Chile, Croatia, India and Rwanda) the levels found in IMAGES based on mens reports of use of IPV are higher than many of the previous surveys that only included womens reports. For example, in Rwanda, data from the DHS survey estimated IPV levels of 30 percent whereas IMAGES estimates prevalence of IPV against women at 39 percent (mens reports) and 41 percent (womens reports). This suggests that the methodology used by IMAGES for both men and women provided for accurate assessments of the magnitude of IPV. Factors associated with mens use of violence were rigid gender attitudes, work stress, experiences of violence in childhood and alcohol use. In Brazil and India (two among the three countries whose surveys included an item on IPV in the last year), younger men were more likely to report use of
PART 2
PART 3
IPV, although this was statistically significant only in the case of Brazil. Men with higher educational levels report the lowest levels of IPV (although only statistically significant in the case of Brazil, Chile and India). Analysis of IPV in the last year presents the same patterns regarding the link with younger and more educated men with a statistically significant association in the case of India. In all other countries, with the exception of Mexico, there is a clear association between men reporting economic stress and higher rates of lifetime use of IPV, with a statistically significant association in Chile, Croatia and India. In Brazil and India, there is also a significant association related to the perpetration of IPV in the last year. Men who hold gender-inequitable attitudes are more likely to report having used IPV, with a clear pattern in all countries (statistically significant in Chile, Croatia, Mexico and Rwanda, and also statistically significant for IPV use in the last year in the case of Brazil). Alcohol abuse was also shown to be associated with IPV, again consistent with previous research. In all research sites, men who regularly (at least monthly) have five or more drinks at one time report higher levels of IPV compared to men who do not report this heavy drinking. This association was statistically significant in Chile and Mexico for lifetime IPV use and in India for IPV perpetration in the last year. IMAGES results confirmed a strong association in all countries between witnessing violence within the household of origin during childhood and IPV during adulthood, which also is consistent with previous research. This is the only variable that presents a statistically significant association in all countries, both during lifetime and in the last 12 months. Mens reports of perpetration of sexual violence against women range from 6 to 29 percent, the majority of this against a stable female partner in the cases of India and Mexico.28,29 Figure 30 shows the percentage of men who report ever having used sexual violence (including acts against a partner and/or other women either alone or in a group),
45
results are difficult to compare with other studies, with the exception of the MRC study in South Africa, whose methodology IMAGES followed. However, the levels found in IMAGES are high and disturbing, and call attention to the need for The percentages of men who reported that they have increased efforts for sexual prevention. ever used sexual violence (against a partner or against any woman) vary from 2 percent (in Brazil) Sexual violence against a stable partner was the most common form reported by men in India and to 25 percent (in India). In the other countries, Brazil. In India, nearly 20 percent of men reported mens lifetime reported use of sexual violence was ever having carried out sexual violence against a around 9 percent, except Mexico, which as in the female partner, with lower but still high rates in case of IPV, also shows lower rates (and far lower other countries. The levels of sexual violence against than those reported by women in Mexico). These the percentage of men who have ever forced their current or ex-partners to have sex and the percentage of men who have forced their current or ex-partners to have sex in the last 12 months.
Brazil 0 1 2 Chile 4 4 9 Croatia 1 4 9 India 14
Sexual Violence against a partner in the last year Sexual Violence against a partner ever Sexual Violence against anyone ever
20 24 Mexico 2 3 4 Rwanda 2 4 9
0 5 10 15 20 25
% of men who report perpetrating form of sexual violence
Figure 30: Mens Reports of Using Sexual Violence
46
PART 1
PART 2
PART 3
Brazil Robbery Fights Own Firearm Arrested Prison 10 22 3 13 3
Chile 22 16 10 29 4
Croatia 36 18 11 18 6
India 4 7 1 6 1
Mexico 3 13 3 28 3
Rwanda 13 5 4 25 16
Table 8: Mens Reports of Criminal Activity, Firearm Ownership and Imprisonment (percent of all male respondents)
a female partner in the last 12 months varied from .05 percent in Brazil to 14 percent in India; in other countries, the figures are around 1 to 4 percent.30 Similar to IPV, men who experienced violence growing up and showed more inequitable attitudes were more likely to report using sexual violence, as were men who reported high rates of alcohol use. In terms of factors associated with mens reports of sexual violence, holding gender-inequitable attitudes and witnessing violence within the household during childhood are the two variables that present a clear pattern in all countries (nearly always statistically significant). As in the case of IPV, men who showed more gender-equitable attitudes were less likely to perpetrate sexual violence (statistically significant everywhere but Brazil). Men who witnessed violence during childhood were more prone to use sexual violence, which was statistically significant in all countries except Brazil, although results there showed a similar pattern. Work stress was found to be significantly related to mens use of sexual violence only in Croatia. Alcohol abuse was significantly associated with mens reports of sexual violence in Chile, Croatia and Mexico. Unlike data from South Africa, where gang rape has been found to represent a significant portion of mens reports of sexual violence (Jewkes, et al. 2009), and where 9 percent of men interviewed
reporting having participated in a gang rape, among IMAGES sites, mens reports of having participated in gang rape were less than 1 percent in Brazil, India, Mexico and Rwanda and less than 2 percent in Chile and Croatia. In terms of factors associated with mens reports of having participated in gang rape, no clear patterns or significant associations were found, except for gender attitudes in the case of Croatia and alcohol abuse in the case of Chile. Relatively high percentages of men report ever having participated in criminal or delinquent acts, and between 6 percent and 29 percent of men report ever having been arrested. Table 8 shows the percentage of men who report ever having been involved in robberies, fights with weapons, firearm ownership and having been arrested and imprisoned. The highest rate of reported participation in robbery was in Croatia (36 percent) and the lowest in India (4 percent) and Mexico (3 percent). Fights with weapons were the highest in Brazil, where one in five men reported this practice, followed closely by Croatia. In India and Rwanda, this was not a common practice, with levels less than 10 percent. Croatia and Chile show the highest rates of firearm ownership, with the lowest rates found in India. The Latin American countries show the highest
47
levels of men who have been arrested (Chile and Mexico) or imprisoned (Brazil)31 followed closely by Rwanda, and with the lowest levels for both of these found in India. Interestingly, India had among the highest rates of mens reports of IPV and sexual violence but low rates of other forms of violence and criminal behavior by men, while Latin America (and Croatia) had lower rates of IPV and sexual violence but high rates of mens involvement in other forms of violence (and high rates of incarceration).
Men are more likely to report having experienced violence outside the home than are women. Womens rates of experiencing violence outside the home in the last year, ranging from 2 percent in Mexico to 15 percent in Rwanda, were less than mens reports of experiencing violence outside the home in the past three months. For men, the percentages for experiencing violence outside the home in the past three months ranged from 5 percent in Brazil to 23 percent in Chile. Although the forms of violence that women experience outside the home In terms of factors associated with mens participation may be different than that experienced by men in criminal activity, firearm ownership and incarcera- (both in kinds of violence, nature of threats, and power differentials between perpetrator and victim) tion, mens socio-economic situation was the most significant. Men with lower educational attainment, and thus such numbers may not be strictly comparable, they are indicative of how widespread for example, were more likely to have reported participating in robbery (with a similar pattern seen in violence is in the lives of men and women, inside most countries and statistically significant in Croatia and outside the home. and Mexico). Similarly, men with lower educational attainment were more likely to report being involved in fights with weapons, with significant associations in Brazil and Croatia. Likewise, in Mexico, 75 percent of men who do not have a formal education have been in prison compared with 8 percent of men with the highest levels of education. In Chile, Croatia and Mexico, gender attitudes were found to be associated with criminal activity, with men adhering to more rigid or inequitable norms more likely to have participated in some forms of criminal behavior. For example, in Croatia, robbery, having been arrested, fighting with a weapon, and having a firearm were associated with gender-inequitable attitudes (although there was only a statistically significant association with the last two factors). Discussion: Violence and Criminal Practices Although womens reports of physical IPV in the same settings are slightly higher than mens (with the exception of Mexico, where they are considerably higher), the results affirm that in a well-designed survey, men will disclose their use of various forms of violence, including IPV and sexual violence at rates that are quite close to the rates that women in the same setting report having experienced. In terms of factors associated with mens reports of IPV, the results from IMAGES confirm previous
Men who owned firearms or carried out other violence or criminal behavior were more likely to report having used IPV. Another interesting aspect to mention is the statistically significant association in all countries of involvement in robbery, fights and being in prison with use of IPV. Men who reported one or more of these behaviors were more likely to also report having used IPV. In addition, men who reported owning firearms in the case of Brazil, Chile, Croatia and Mexico were more likely to have used IPV (with statistically significant associations in the Brazil, Croatia and Mexico).
48
Jan Sandberg/Influential Men
PART 1
research showing associations between mens use of IPV and their childhood experiences of violence, holding inequitable attitudes about gender, workrelated stress and alcohol use. Furthermore, the strong association between mens reports of economic stress and lifetime use of IPV confirm the need to understand and pay greater attention to work, poverty and economic marginalization and their relationships to mens practices as well as to womens vulnerability to IPV. The finding of the association between having carried out other forms of violence and mens use of IPV affirm other research suggesting that IPV and other forms of criminal behavior form part of a broader gendered culture of violence. Mens high rates of incarceration or confrontations with the police, as well as high rates of experiencing violence outside the home (in childhood and adulthood) confirm how much violence is a common part of too many mens and womens lives and how such forms of violence interact.32
PART 2
PART 3
Transactional Sex Globally, numerous studies have affirmed how gender-related norms and social expectations about mens sexuality encourage mens purchasing of sex (Je-Cannon 2006; Ricardo and Barker, 2008). A recent review of the literature on men, masculinities and transactional sex suggests that the global percentage of men who purchased sex in the last 12 months was estimated at 9 to 10 percent, with estimates from 13 percent to 15 percent in the Central African region, 10 to 11 percent in Eastern and Southern Africa, and 5 to 7 percent in Asia and Latin America (Carael et al, 2006). Other available data indicates, however, that the percentage of men who buy sex varies widely among countries and cultures, ranging from a few percent in some countries to 40 percent in others (Joe-Cannon, 2006; IOM, 2003). What do men in the IMAGES sites report as their practices related to purchasing sex? What factors are associated with mens purchasing of sex? Key IMAGES questions looked at mens lifetime reports of paying for sex, age of sex workers for whom they paid for sex, whether they thought the sex worker might have been coerced or trafficked, and mens attitudes about transactional sex or sex work (and sex workers). We asked men about paying for sex and other forms of transactional sex such as giving gifts in exchange for sex.
Have you ever had sex with any sex worker? %

Brazil (n=626) Chile (n=1153) Croatia (n=1458) India (n=1527) Mexico (n=994) Rwanda (n=2231)
Have you ever had sex with a female sex worker? % 54 22 11 24 18 15
Have you ever had sex with a male sex worker? % 1 1 0 2
Have you ever had sex with a transvestite sex worker? % 2 0 0 1
56 23 12 27 18 16
Table 9: Men's Reports of Transactional Sex
49
Brazil 56 17 8 Chile 22 14 7 Croatia 12 5 8 India 27 48 34 Mexico 18 14 7 Rwanda 16

0 10 20 30 40 50 60 Has had sex with any sex worker OF THOSE (blue bar) suspected that sex worker was <18 OF THOSE (blue bar), suspected that sex worker was forced or sold into prostitution
% reporting transactional sex experience
Figure 31: Transactional Sex Experiences: Mens Reports that a Sex Worker with Whom they Had Sex was Under 18 or Forced/Sold into Prostitution Between 16 percent and 56 percent of men in the countries surveyed say they have paid for sex at least once. IMAGES results are consistent with other surveys on mens reports of paying for sex: between 16 percent of men (in the case of Rwanda) and up to 56 percent of men (in the case of Brazil) reported that they had ever paid for sex with any sex worker (male, female or transvestite, although the vast majority of those reported sex with female sex workers). Sex with male sex workers ranged from a high of 1.8 percent in India, to roughly 0.5 percent in the case of Brazil, Croatia and India, with negligible rates in the case of Rwanda. Mens reports of having paid for sex with transvestite sex workers ranged from greater than 1 percent in Brazil and India and less than 1 percent in Chile, Croatia and Rwanda. Men with lower educational attainment and less genderequitable attitudes are more likely to have paid for sex. Men with lower educational attainment were more likely to report having paid for sex in the case of Croatia and India. In Brazil, a more complex association emerged in terms of having paid for sex and educational attainment; men with the lowest and the highest levels of education reported higher rates of paying for sex. 33 In all countries except Brazil, the men who most support rigid, inequitable gender norms (as assessed by the GEM Scale) were more likely to report having had sex with sex workers. In Brazil, the association was not statistically significant but there was an association in the opposite direction: men with more gender-equitable attitudes were more likely to report having paid
50
PART 1
PART 2
PART 3
for sex. In the case of Croatia and India, men who IMAGES also asked men who reported having paid reported work-related stress were more likely to report for sex if they thought those sex workers were under having paid for sex. the age of 18 and also asked if they thought that a sex worker with whom they had sex (regardless of age) Men who are unsatisfied in their sexual relationship was coerced into being a sex worker. In many cases, with a stable partner are more likely to report having men do not know with certainty about either of these paid for sex. points, but it is nonetheless useful to understand In the cases of Chile, Croatia and Mexico, men who their assessments of both. Of all men who report reported being unsatisfied with their current sexual having sex with a sex worker, significant proportions relationship with their stable partners were more (ranging from 48 percent in India to 5 percent in than two times more likely to report having had sex Croatia) suspected that in at least once incidence, a with a sex worker than men who reported being sex worker with whom they had sex was younger than sexually satisfied with their current stable partner. age 18. Of this same group, anywhere from 7 percent This association was also seen in Rwanda and (in Chile) to 34 percent (in India) suspected that in at Brazil, although it was not statistically significant. least one incidence a sex worker with whom they had sex had been forced or sold into prostitution (Figure An important minority of men who have paid for sex 31). Looking specifically at India and Brazil, this sughave done so with a sex worker they think was under gests that 4 percent of Brazilian men surveyed and 9 age 18 or who was coerced into sex work. percent of Indian men surveyed report ever having
Brazil 37 82 Chile 57 62 91 Croatia 58 44 81 India 84 70 65 Mexico 21 3 44

0 10 20 30 40 50 60 70 An adult woman working as a sex worker is wrong morally Sex work violates an adult womans rights Working as a sex worker is an adult womans own choice 80 90 100
% who agree
Figure 32: Mens Attitudes about Adult Female Sex Workers
51
Brazil 50 85 Chile 68 63 92 Croatia 67 42 78 India 84 70 69 Mexico 30 2 47

0 10 20 30 40 50 60 70 80 An adult man working as a sex worker is wrong morally Sex work violates an adult mans rights Working as a sex worker is an adult mans own choice 90 100
% who agree
Figure 33: Mens Attitudes about Adult Male Sex Workers
had sex with a sex worker they thought was forced or sold into prostitution. Men have contradicting attitudes about sex work: many think it is both morally wrong and violates the rights of those involved but many think it is also a womans choice to be a sex worker, even if she is younger than 18. Across the countries men showed contradictory attitudes about sex work. On the one hand, between 65 percent and 91 percent of men believe that it is a womans choice to be a sex worker (with 44 percent in the case of Mexico). At the same time, between 21 percent and 84 percent think that sex work is morally wrong. High percentages of men thought that sex work violates a womans rights while at the same time saying it was her choice (Figure 32). Men who think sex work is morally wrong were less likely to have paid for sex in the case of Brazil, Chile, Croatia and Mexico. Men who
believe it is a womans choice to be a sex worker were more likely to have paid for sex in the case of Croatia. Men had fairly similar attitudes about adult male sex workers, although the percentage who thought male sex work was morally wrong was slightly higher in a few settings (Figure 33). In addition, men showed similarly contradictory attitudes about sex work with girls younger than 18 (which is illegal in all the countries studied), with relatively high rates of men thinking it is a girls choice to be sex worker or trade sex for goods or money, while only small percentages think that this practice by girls is morally acceptable (Figure 34). Figure 35 shows the responses in terms of mens attitudes about sex work carried out by boys younger than age 18. Interestingly, significantly more men thought it was a boys choice to be engaged in sex work than it was for girls.
52
PART 1
Discussion: Transactional Sex Between one out of five and more than half of men in the cities surveyed reported that they had ever paid for sex, confirming that transactional sex is a common practice. Men with lower levels of education, men who were less satisfied in their sexual relationships with their stable partner and men with the most gender-inequitable views are generally those most likely to have paid for sex. Furthermore, a small but sizable minority of men in Brazil and India affirm that they have had sex with sex workers younger than 18, and with a sex worker they think was coerced into sex work. When asked what they think about sex work and sex workers, men showed convergent and in many ways conflicting views. At the risk of oversimplifying mens responses on these issues, they seem to be saying that sex work violates womens rights, that it is morally wrong, but that it is a womans choice (even if she is younger than 18). Reducing or questioning mens practice of paying for sex has seldom been part of large-scale efforts around engaging men. While the ethical and policy issues involved are complex often centered around whether sex work is inherently exploitative these
PART 2
PART 3
results affirm that it is a widespread practice related to salient versions of manhood, and that men hold views about it that are as complex (and contradictory) as the policy debates currently underway.
Brazil 57 27 Chile 43 18 Croatia 29 6 India 63 16 Mexico 59 4

0 10 20 30 40 50 60 I see nothing wrong with a woman <18 working as a sex worker Sex work is a woman <18s own choice
% who agree
Figure 34: Mens Attitudes about Female Sex Workers Younger than 18
Robyne Hayes/ICRW
53
Brazil 82 27 Chile 88 17 Croatia 89 7 India 86 18 Mexico 59 4

0 10 20 30 40 50 60 70 I see nothing wrong with a boy <18 working as a sex worker Sex work is a boy <18 own choice 80 90
% who agree
Figure 35: Mens Attitudes about Male Sex Workers Younger than 18
54
David Snyder/ICRW
PART 1
PART 2
PART 3
Brazil 88 64 66 65 Chile 90 74 72 71 Croatia 90 43 35 48 India 47 77 80 71 Mexico 87 41 39 39

0 10 20 30 40 50 Supports womens rights: Men dont lose out Supports quota: Fixed proportion of places in govt. Supports quota: Fixed places in universities Supports quota: Fixed proportion of executive positions 60 70 80 90 100
% who support
Figure 36: Mens Support for Gender Equality Overall and Specific Quota-Based Policies
Knowledge and Attitudes about Policies and Laws Promoting Gender Equality In the past 10 years, there has been increased attention to engaging men in gender equality at the policy level. Notably, in all of the countries included in IMAGES, there are now laws that stipulate sanctions for men convicted of using physical violence against intimate partners. What do men think about these policies and laws related to gender-based violence, as well as policies related to gender equality in general? What percentage of men
have been exposed to messages or campaigns (or report having seen them), focused on gender-based violence, reducing homophobia, mens involvement in caregiving or as fathers? In terms of gender equality overall, men in all the countries, with the notable exception of India, are generally supportive of gender equality, with 87 percent to 90 percent saying that men do not lose out when womens rights are promoted (Figure 36). Even when asked about specific policies, such as quotas for
55
women in executive positions, university enrollment or government, mens support for such policies is reasonably high, with roughly 40 percent to 74 percent of men supporting such quotas. In India, the pattern was the opposite: only 47 percent of men of Indian men supported gender equality overall, while higher percentages (71 to 80 percent) supported quotas for women in work, education and government. Gender-based violence is the policy issue that has received the most attention in terms of engaging men in gender equality, and one that many men have heard about, in all the countries included in IMAGES. Figure 37 shows that 41 percent to 85 percent have ever seen an advertisement related
to GBV, while 17 percent to 77 percent have ever seen a campaign related to GBV. Figure 38 shows that in addition to being exposed to campaigns about VAW, men are more and more aware of laws against VAW. Between 88 percent to 96 percent of men know about the laws related to VAW in their countries. In all countries but India this high awareness of VAW laws coincides with low levels of agreement with the statement There are times when a woman deserves to be beaten, suggesting that (with the exception of India), men in IMAGES study countries have both the knowledge and attitude that might lead to decreased perpetration of intimate partner violence. As the third bar in Figure 38 asserts,
Brazil 18 41 4 Chile 49
Ever encountered campaign questioning VAW Ever seen an advertisement questioning VAW Ever participated in an activity questioning other mens use of VAW
84 13 Croatia 77 85 10 India 54 78 13 Mexico 47 82 8

0 10 20 30 40 50 60 70 80 90 100
% who have encountered campaign/advertisement or participated in activity
Figure 37: Mens Exposure to Messages and Campaigns About Violence Against Women
56
PART 1
PART 2
PART 3
Brazil 95 19 24 Chile 10 29 Croatia 97 12 33 India 92 65 37 Mexico 93 6 17 Rwanda 96 21 39

0 10 20 30 40 50 60 70 80 90 100 Knows of VAW law in country Agrees: There are times when a woman deserves to be beaten Has ever perpetrated physical violence against partner
88
% in category
Figure 38: Knowledge About VAW Law, Attitudes about VAW, and Perpetration of Violence
however, rates of lifetime perpetration of IPV remain high in all IMAGES countries. While men are clearly receiving messages about GBV and know about the existence of laws related to GBV, Figure 39 confirms the contradictory attitudes that men have about the existing laws related to GBV. Between 39 percent and 92 percent of men think the laws on GBV in their country make it too easy to bring charges against men. Given the relatively low number of men who have been formally charged under these laws in all the countries included, this is hardly an accurate assessment of these laws. At the same time, between 40 percent
and 88 percent of men in the countries surveyed think these laws do not offer enough protection to women who experience GBV. In terms of other gender-related campaigns or activities, between 12 percent and 23 percent of men have seen campaigns promoting mens involvement as fathers and 6 percent to 11 percent say they have ever participated in an educational or discussion session about their roles as fathers (Figure 40). In addition, 10 percent to 38 percent of men report ever having seen a message in their country about reducing homophobia or messages or campaigns promoting the rights of same-sex attracted individuals.
57
Discussion: Knowledge and Attitudes about Policies and Laws Promoting Gender Equality In general, the impression that emerges is that GBV has been the locus of most large-scale campaigns targeting men in the area of gender equality, but that men continue to have conflicting attitudes about these laws, may not understand the laws and may feel defensive about them. Furthermore, it appears that the combination of increased awareness of GBVrelated laws and high levels of attitudes rejecting violence (in all countries but India) has not coincided with a concomitant reduction in levels of intimate partner violence. This report does not test the statistical association between these factors, but nonetheless these findings suggest the need for
government and civil society to carry out more nuanced, long-term and wider reaching public education particular outreach to men about these laws. For example, campaigns should go beyond the simplistic violence against women is against the law, which still prevail in some parts of the world. We did not ask men about their impressions of messages related to their roles as fathers, but we can reasonably assume that they have fewer negative reactions to them. These findings suggest the need for national and local campaigns that engage men on multiple topics, including some that men perceive as positive and reflective of their perceived needs, challenges, and health-related and other vulnerabilities.
Law makes it too easy to charge men
Law doesnt offer enough protection for victims
Brazil 80 72 Chile 76 88 Croatia 39 85 India 88 81 Mexico 81 66 Rwanda 92 40

0 10 20 30 40 50 60 70 80 90 100
% of men who agree or partially agree
Figure 39: Mens Attitude about Laws Criminalizing Intimate Partner Violence
58
PART 1
PART 2
PART 3
Brazil 15 8 10 Chile 16 8 18 Croatia 23 6
Encountered campaign promoting fathers involvement Participated in activity to promote fathers involvement Encountered campaign questioning homophobia
38 India 12 12 15 Mexico 18 8 24
0 10 20 30 40
% of men who responded yes
Figure 40: Mens Exposure to Messages and Campaigns Related to Fatherhood and Related to Sexual Diversity
Shana Pereira/ICRW
59
PART III: Conclusions

Are men evolving in their practices and attitudes related to gender equality? We can affirm in reviewing these initial comparative findings from IMAGES that change seems to be happening. Overall, these results attest that gender equality can be and should be seen as a gain for women and men. Policies, programs and public education campaigns aimed at engaging men must move creatively and boldly beyond small-scale approaches to capitalize on this apparent emerging shift in mens attitudes. That said, we must seek to understand the dimensions of work-related stress and understand that income and employment are key elements in mens (and womens) identities, self-esteem, sense of purpose in life and social relations.
Gender attitudes matter. Mens attitudes about gender whether they believe in a set of norms related to men and women being equal were consistently associated with their practices across the vast majority of topics examined. Ultimately, the harmful practices of men are rooted These results affirm the need to change structural in gender norms, economic and work-related stress, issues campaigns to involve men in childbirth mens childhood experiences of violence and the have proven very successful in Chile, for instance rigid class differences that lead to the arrest and while still working to deconstruct the inequitable imprisonment of sizable minorities of men. If we norms that many men and boys continue to interseek to end violence, and to achieve gender equality, nalize related to their use of violence, limited we need to understand the conditions that lead to participation in care work, and limited help- and these practices. health-seeking behavior. The following key findings from IMAGES can help us more effectively engage men in achieving gender equality:
Men are generally positive about gender equality. In all the countries, with the possible exception of men in India, men did not see gender equality as a zero-sum game where gains for women mean Work-related stress is a major factor and common losses for men. Public education campaigns and occurrence in mens lives. policies should do more to emphasize to men that Between 34 and 88 percent of men interviewed say achieving gender equality benefits all societies, that they are frequently stressed or depressed children, women and men themselves. It is clear because of not having enough work or income. from some women and men interviewed that sizeTen to 50 percent of unemployed men say they have able portions of men are practicing and living at considered leaving their families because of workleast some elements of gender equality. Efforts to related stress. Mens stress related to not having promote gender equality should tap into the support enough income or work and achieving the role of that some men already show for gender equality and provider was a key factor associated with perpetra- womens empowerment, despite the apparent contion of violence, higher rates of incarceration, tradictions in mens responses that support gender higher rates of alcohol abuse and higher rates of equality in the abstract while resisting it in practice. suicidal thoughts. These initial findings suggest that Educated men are more likely to live gender measuring work-related stress, rather than simply equality in their daily lives. measuring income, may be a useful way to understand how income and employment (and Education is an important factor contributing to unemployment) affect mens and womens lives. mens more gender-equitable attitudes and Achieving gender equality requires overcoming the practices. Using the education system, particularly notion that mens only social role is as provider. secondary education, can be an important locus for
60
PART 1
PART 2
PART 3
school and community); reduce alcohol abuse; protect and support survivors of all forms of violence; and rehabilitate rather than solely punish perpetrators of violence. Related to this, the IMAGES data also affirm the complex cultural and social transChildhood experiences matter, affirming the need mission of violence. The relatively high percentage to engage men in gender-equitable practices in of men with low educational levels interviewed who the lives of children. have been arrested or imprisoned is alarming and Men who report that their fathers were involved in must be discussed as a social justice issue that has child care or domestic work were significantly more gendered components. These data affirm that far likely to be involved themselves. This affirms the too many men have experienced violence, importance of mens interactions with their chilperpetrate violence (in multiple forms) and have dren in both promoting child well-being and as a been arrested or imprisoned with limited or no matter of equity in their families. Mens participaaccess to rehabilitative programs or support servtion in traditionally feminine activities like care ices, and that this violence is directly related to the work is a powerful political act as well; men who violence that many men perpetrate against women. are involved in care work demonstrate to the next Address the structural factors of gender equality. generation of boys and girls, both in the home and in their communities, that care work should be Engaging men in gender equality requires being equally shared between women and men. empathetic with mens lived experiences the unspoken depression, suicidal thoughts, high levels A sizable portion of men have experienced and of childhood experiences of violence, and their high used violence in multiple ways. levels of work-related stress. But being empathetic IMAGES results affirm the need for integrated toward the structural conditions of mens lives is not to make excuses for the violent and oppressive pracviolence prevention interventions and policies that tices of some men. It is, instead, an affirmation of the seek to change gender-related attitudes associated with violence; create spaces where men can discuss need to move beyond a superficial understanding of gender equality toward addressing the structural and overcome violence they have experienced; but changeable factors that underpin it. reduce violence against children (in the home,
doing even more with girls and boys to promote gender equality. There also is some evidence of generational change as well as the role that co-habitation plays in shaping views about gender equality.
Robyne Hayes/ICRW
61
References
Bannon, Ian, and Maria C. Correia. 2006. The Other Half of Gender: Mens Issues in Development. World Bank Publications, June Barker G, Ricardo C. 2005. Young men and the construction of masculinity in sub-Saharan Africa: implications for HIV/AIDS,conflict and violence. Washington, DC, World Bank. Emslie, Carol, Damien Ridge, Sue Ziebland, and Kate Hunt. 2006. Mens accounts of depression: Reconstructing or resisting hegemonic masculinity? Social Science & Medicine 62, no. 9 (May): 2246-2257. doi:10.1016/j.socscimed.2005.10.017. Garca-Moreno, C.; Jansen, H.; Ellsberg, M.; Heise, L.; Watts, C. 2005. WHO Multi-country study on Womens Health and Domestic Violence against Women. WHO: Switzerland.
Brown, J. & Chevannes, B. 1998. Why man stay so: An examination of gender socialization in the Caribbean. Gupta, Geeta Rao. 2000. Gender, Sexuality, and University of the West Indies, Kingston, Jamaica. HIV/AIDS: The What, the Why, and the How. Budlender, Debbie. 2008. The Statistical Evidence on Plenary Address XIIIth International AIDS Care and Non-Care Work across Six Countries. United Conference, Durban, South Africa. Washington, DC: Nations Research Institute for Social Development. International Center for Research on Women. Geneva. International Center for Research on Women Carael, M, E Slaymaker, R Lyerla, and S Sarkar. 2006. (ICRW). 2008. Seven Priorities, Seven Years to Go: Progress on Achieving Gender Equality Caren Grown, Clients of sex workers in different regions of the Geeta Rao Gupta, Aslihan Kes. ICRW, Washington, DC. world: hard to count. Sexually Transmitted Infections 82 Suppl 3 (June 2006): iii26-33. International Organization for Migration (IOM). 2003. Is Trafficking in Human Beings Demand Connell. R.W. 1994. Masculinities. Berkeley, CA. Driven? A Multi-Country Pilot Study, December University of California Press. 2003, Migration Research Series, No. 15, available at: http://www.unhcr.org/refworld/docid/415c6fa34.html Connell, R. W. 2000. The men and the boys. Cambridge, UK: Polity Press. Jewkes, R., Sen, P., Garca-Moreno, C. 2002. Sexual Courtenay, Will H. 2000. Constructions of masculin- violence. In: E. G. Krug et al. (Eds.) World report on ity and their influence on mens well-being: a theory violence and health. Geneva, Switzerland: World of gender and health. Social Science & Medicine 50, Health Organization. no. 10 (May 16): 1385-1401. doi:10.1016/S0277Joe-Cannon, I. 2006. Primer on the male demand for 9536(99)00390-1. prostitution. N. Amherst, USA: The Coalition Draper, J. 2002. It was a real good show. The ultra- Against Trafficking in Women. Available at: http://action.web.ca/home/catw/attach/PRIMER.pdf sound scan, fathers and the power of visual knowledge. Sociology of health and illness, 24(6), 771-795. Kaufman M. 1993. Cracking the armour: power, pain and the lives of men. Toronto, Viking. Ekelin, M., Crang-Svalenius, E., Dykes, A-K. 2004. A qualitative study of mothers and fathers Kimmel M. 2000. The gendered society. Oxford, experiences of routine ultrasound examination in Oxford University Press. Sweden. Journal of Midwifery, 20, pp.335-344.
62
REFERENCES
Marsiglio W. 1988. Adolescent male sexuality and heterosexual masculinity: a conceptual model and review. Journal of Adolescent Research, 3:285303. Moller-Leimkuhler, Anne Maria. 2003. The gender gap in suicide and premature death or: why are men so vulnerable? European Archives of Psychiatry and Clinical Neuroscience 253, no. 1 (2): 1-8. doi:10.1007/s00406-003-0397-6. National Center on Fathers and Families. 2002. The Fathering Indicators Framework: A Tool for Quantitative and Qualitative Analysis. Philadelphia, PA; Pulerwitz J, Barker G. 2008. Measuring attitudes towards gender norms among young men in Brazil: development and psychometric evaluation of the GEM Scale. Men and Masculinities. Pyne, Hnin Hnin, Mariam Claeson, and Maria Correia. 2002. Gender Dimensions of Alcohol Consumption and Alcohol-Related Problems in Latin America and the Caribbean. World Bank Discussion Paper 433, World Bank, Washington, DC. Ricardo, C.; Barker, G. 2008. Men, masculinities, sexual exploitation and sexual violence. A Literature Review and Call for Action. Rio de Janeiro, Brazil: Instituto Promundo, MenEngage. Rivers K, Aggleton P. 1998. Men and the HIV epidemic, gender and the HIV epidemic. New York, HIV and Development Programme, United Nations Development Programme. WHO (World Health Organization). 2004. World Health Report. Geneva: 2004. WHO. 2007. Engaging men and boys in changing gender-based inequity in health: Evidence from programme interventions. Geneva. Available at: http://www.who.int/gender/documents/Engaging_ men_boys.pdf. The World Bank. 2005. World Development Indicators. Washington, DC: World Bank. The World Bank. 2007. Gender Equality as Smart Economics. A World Bank Group Gender Action Plan. September 2007.
63
Annex I: Overview of Related Studies

Key Findings from Norway Study on Gender Equality and Quality of Life
Summary by study co-author ystein Gullvg the household level had been achieved, but the difHolter, Center for Gender Studies, University of Oslo ference was not dramatic. To be sure, some household tasks remain feminine in the percepThe 2007 Gender Equality and Quality of Life survey tions of respondents, but some have shifted from in Norway provided much of the conceptual feminine to neutral (or gender-balanced) as background and many of the questions for IMAGES. compared to data from 1988. For example, cleaning The Norway study was groundbreaking in several the house remained feminine while shopping and respects. It applied a multidimensional gender equality making food were neutral. Caring for children is model where the dimensions (including normative more gender-balanced than 20 years ago as well. gender equality, practical gender equality and resource Sixty percent of fathers reported that they were balance in couples) were measured by wide sets of equally involved in the daily care of children. variables. It also included more detail than earlier gender surveys, incorporated a life-course perspective and Responsibility for and organization of household tasks maintained a practical focus that went beyond opinoften remain more traditionally gendered than the ion-based surveys to include self-reported practices in performance of those tasks themselves. Although men terms of division of household tasks. The 16-page played a greater role in traditionally feminine tasks, questionnaire was answered by 2,805 men and women few of these households reported that these tasks had with a response rate of 41 percent. become primarily the mans responsibility. Even if the man and woman shared traditionally feminine The Norwegian survey was implemented in a coun- tasks more equally, it was still most often the woman try with a relatively high level of gender equality, at who had the main responsibility and was seen as the least on the normative level researchers would organizer of these tasks. This was true of domestic expect to find gender equality (Norway is always duties and responsibilities in the so-called reproducnear the top of the World Economic Forums annual tive sphere. Conversely, production sphere tasks Global Gender Gap Index, for instance). Also, Norremained the mans responsibility, including way is a rich country with fairly moderate class household repairs and projects, as well as income differences and a well-functioning welfare system. generation and financial provision for the family. A All of these macro-level conditions mean that direct common rule emerged: the greater the social orientacomparisons between Norway and the different tion of a task, the more feminine; the more technical countries in IMAGES should be done carefully. This orientation of the task, the more masculine. summary highlights the key results of the Norway study although some of them are not immediately Norways gendered professional environment remained comparable with IMAGES findings. unchanged between 1988 and 2007 as compared to these measurable changes in households. Despite Main results included the following: increasing gender equality in many areas, including The proportion of households where domestic tasks political representation and educational achievement, and care work are distributed equally between men the economic imbalance between women and men and women has risen significantly when compared remained, as other research has shown. The gendered with 1988 data. For example, 75 percent of men wage gap was not reduced in Norway between 1988 reported that the woman prepared food in 1988, and 2007. Furthermore, only 50 percent of men compared to 50 percent in 2007. Here, as in other reported gender-equal cooperation in their jobs in studies, women were more skeptical that equality at 2007, the same as in 1988.
64
ANNEXES
Norways new paternity leave or father quota has produced broad positive changes in household and parenting dynamics. Taking parental leave after the birth of a child led to more subsequent responsibility for daily childcare by men and more shared parental roles overall. Young mothers were more satisfied with their relationships and their quality of life if their partners had used the quota. On the whole, a majority of women and men supported an extension of the father quota; the quota has since increased from four weeks in 2007 to 10 weeks in 2010. Survey results suggest that gender equality may be more experience-driven and less ideologicallydriven than is commonly recognized. Gender equality is a strong norm in todays Norway. As many as 90 percent of survey respondents shared the opinion that men and women should share three central tasks equally: care for children, housework and wage work. Although gender equality norms do influence the daily implementation of these tasks, material and practical matters play a larger role than is commonly recognized. These include the income balance and other economic resources on the couple, as well as experiential factors from ones childhood, ones psychological gender formation, the gender culture in local context and ones professional experience. These influences together, rather than gender equality norms or political ideology alone, determine whether an individuals or couples decisions, choices and practices are in fact gender equal. Gender equality benefits both women and men. The positive effects of gender equality for both women and men were most evident in the realm of family and household dynamics. Those holding more gender-equitable attitudes and reporting a more gender-equitable division of tasks and decision-making reported increased quality of life and relationship satisfaction (both women and men). Gender equality at home also was consistently associated, among men and women, with a reduction in couple conflict and chance of divorce, as well as with a reduction of violence including lower violence against children. The latter finding that gender equality leads to a reduction of intimate partner violence and violence against children was one of the strongest and most significant in the survey. Genderequal men and women also got a relatively good bill of health regarding depression, suicide and health. Do these findings mean that Norwegians now have a clear path forward to improved well-being and gender equality? The answer is not completely, but new insights and possibilities have certainly come to light. As was the case in the IMAGES study countries, and despite the success of their new paternity leave quota, Norwegians remain divided (with men the most skeptical) on gender quotas and gender equality policies. Equality in the home enjoys widespread support, but active state regulation less so. These opinions exist alongside findings from the survey that gender equality precipitates positive effects on the health and wellbeing of women and men. The Gender Equality and Quality of Life survey significantly advanced empirical knowledge of Norwegian mens and womens attitudes and behaviors related to gender equality. In borrowing and adapting elements of the survey for application internationally, IMAGES has further advanced the social benefit of this project and expanded knowledge about gender equality attitudes and behaviors on a much larger scale. Much more work remains to be done, and qualitative research insights will be crucial to understanding the nuances of these quantitative findings.
This summary is based on the report: Holter, ystein Gullvg; Svare, Helge & Egeland, Cathrine 2009: Gender Equality and Quality of Life A Nordic Perspective. Nordic Gender Institute and Work Research Institute, Oslo. See: http://www.nikk.no/?module=Articles;action=Article.publicShow;ID=892
65
Key Findings from South Africa Study on Men, Masculinities, Violence and HIV
Summary authored by Rachel Jewkes, Yandisa Sikweyiya, Robert Morrell, Kristin Dunkle, Gender and Health Research Unit, Medical Research Council, Pretoria Introduction and Methods This summary presents some key findings from the 2008 South African study on men, HIV and violence, which was conducted as a partner study to IMAGES, incorporating some items from the Norwegian study, and some that were later used in the IMAGES questionnaire. The South African survey was conducted in a country with a legendary reputation for crime and violence, and the worlds largest HIV epidemic. These problems have been laid at the door of South Africas men. Still scarred by the legacy of apartheid, its education system is dysfunctional, unemployment rife and the wealth distribution is extremely inequitable. This survey sought to better understand men and masculinities and gender relations. The South African study was conducted in three districts in the Eastern Cape and KwaZulu Natal Provinces. It was cross-sectional with a two-stage random sample, with interviews conducted in 215 clusters. Eligible men ages 18-49 were chosen at random in the selected households. Most men interviewed were Black African (85 percent); 9 percent were Indian and 6 percent were colored (mixed race) or White. South Africa has a largely youthful population and reflecting this, half of the men were under 25 years of age and 70 percent were under 30. Respondents used an electronic device, APDAs (Audio-enhanced Personal Digital Assistants), to answer questions that were accessed visually or audibly via the APDA. HIV testing was conducted. The survey interviewed 1,738 men with a response rate of 75.6 percent. Main results include the following: Low educational attainment, high unemployment and low rates of marriage/cohabiting The great majority of men (65 percent) were unmarried and not cohabiting. Educational attainment was poor, with nearly two-thirds not having finished school. Unemployment was very high. Fewer than a third (30 percent) were regularly employed, a quarter had never worked and the others worked seasonally or occasionally. Thus few men were occupying a provider role, which is widely cited as very important in traditional ideas of South African manhood. Gender equity The survey confirmed the patriarchal nature of South African society, with 85 percent of African men agreeing that to be a man it is important to be tough. Nearly two-thirds (60 percent) said that a man should have the final say on decisions at home, and half said that women sometimes deserve to be beaten. Only 10 percent of men scored in the most gender equitable tertile of the Gender Equitable Men (GEM) scale and just under 30 percent were in the lowest equity category. With these scores, South African men appeared more gender inequitable than those from any other IMAGES country. The mens GEM Scale scores were strongly associated with education, income and race. Multivariable analysis showed that while men earning over R2000 per month (about $300) had significantly higher scores than those not earning, the most important factor associated with gender equity was education, with incrementally more education resulting in more gender-equitable attitudes. If attitudes of men of different ages can be seen as indicative of change, there was precious little evidence of this as younger men held attitudes that were no different from those of older men.
66
ANNEXES
Communication with their partner The survey asked about communication between partners about problems or challenges faced at home or work, in the family or community and money. Findings suggest that communication between partners is often poor. For example, among those married or cohabiting, only half (53 percent) had discussed problems in the previous fortnight; this did not vary significantly between racial groups. Many men had not discussed any problems or challenges with their partner in the previous month, including 18 percent of married men and 22 percent of men who were dating. South African mens communication with their partners appeared to be less satisfactory than that of men from most other IMAGES countries, which is characteristic of more inequitable relationships. toward a partner and 28 percent had ever raped a woman or girl. Further, many men had been involved in other crime and violence. A third (34 percent) had stolen or been part of a robbery on more than two occasions, and over a quarter had been involved in a fight with knives (27 percent). Many men had other weapons (20 percent) and a greater proportion of men (11 percent) had ever had possession of an illegal gun. While having been arrested and imprisoned were common experiences (24 percent and 8 percent prevalence, respectively), when compared to other IMAGES countries, South African men were not particularly unusual. This reflects the considerable impunity with which men may engage in crime and violence in the country.
Conclusions Although most of the men in this study had spent Engagement with child care all of their adult lives and much of their childhood Given how many men did not discuss problems in the democratic (post-apartheid era) South Africa, with their partners, it was unsurprising that over a there was precious little evidence that they enjoyed third never discussed personal matters with their material fruits of freedom, or had generally imbibed children. When men lived with their children, many attitudes supportive of gender equity which are men did participate in child care, with half of the found in the countrys constitution. The men men saying they often played with children, helped generally held patriarchal views and many had been with homework or fixed food for them. However, if involved in violence and crime. This did not prevent men did not live with their children, they commonly some of them from being involved with child care, failed to perceive themselves as obligated to provide but if men did not live with their children, the financial or social support, irrespective of ability. norm was not to regularly financially support them. Only half of men who did not live with their children Gender equity in South Africa is still a work in and were earning an income provided monthly progress, but the evidence suggests that change in maintenance. This problem is well recognized in mens attitudes and practices is in many ways the country with very high levels of maintenance predicated upon broader social development, default. In 2002, one urban district in Durban particularly improvements in education. recorded 372,000 complaints for default. Men and violence A very high proportion of men disclosed perpetration of gender-based violence, many more than in most IMAGES countries. Eleven percent of men had been physically violent towards a woman partner in the last year, 42 percent had ever been violent
This study was funded by the U.K. Department for International Development (DFID), and the grant was managed by their local partner Human Life Sciences Partnership (HLSP). Further details of methods and findings related to rape and HIV are available on: http://www.mrc.ac.za/gender/interfaceofrape&hivsarpt.pdf
67
Annex II: Original Conceptual Framework

The graphic below represents the conceptual framework which informed the selection of questions in the IMAGES questionnaires. This report has explored some of the relationships implied in the framework below; additional studies will include multivariate analyses of these and other relationships.
Mens Practices
Violence
Physical violence against partner Perpetration of sexual violence Involvement in criminal acts
Structural
Socio-demographic and cultural factors Economic empowerment Household composition
Relationship dynamics Exposure to policies and campaigns

Fatherhood VAW Sexual diversity Decision-making dynamics Participation in household tasks Couple Communication Support abortion decision
Childhood experiences
Men abused during childhood Childhood household gender dynamics Culture of and experiences of violence at school and neighborhood
Health
Sexuality SRH Care Help and health-seeking behaviors Drug and alcohol abuse
Gender Attitudes
GEM Scale Gender attitudes related to women's rights and economic empowerment
Women and children

Marital status Length of relationship Partners employment situation Number and age of children Previous unions
Parenting
Participation in maternal health Taking paternity leave Caring for children Responsibilities with biological children
Transactional Sex
Sexual relations with sex workers Transactional sex for goods
68
ANNEXES
Annex III: GEM Scale Scores Trichotomized by Country

The figure below presents the distribution of men in each country according to the trichotomized GEM Scale results. As described earlier, this scale was constructed independently for each country based on Cronbach Alpha tests of internal consistency. The final scales included from eleven to fifteen items based on the country (see Table 4). The scale results were trichotomized by their total score. That is, the range of possible scores for that country was divided into equal thirds; thereafter the men scoring in the lowest third (of the possible score, not of the weighted pool of responses) were placed in the low equity category, those in the middle third in the moderate equity category and those in the highest third in the high equity category. Because each countrys scale included different items, they are not strictly comparable with one another. Thus the high proportion of men in the high equity category in any country does not necessarily imply that men in that country are more gender-equitable than those in another, or vice versa. These distributions merely illustrate where mens GEM Scale scores landed among the range of possible answers on that countrys finalized, internally consistent scale.
Brazil 9 33 58 Chile 3 27 70 Croatia 3 15
Low Equity Moderate Equity High Equity
82 India 24 59 17 Mexico 4 22 74 Rwanda 17 53 30

0 10 20 30 40 50 60 70 80 90
Percentage in GEM group
GEM Scale Trichotomized Distribution by Country
69
Annex IV: Sampling Methodologies

This annex describes the sampling strategies and fieldwork methodologies for all six study countries. A brief overview of these methodologies can be found in Table 1. Brazil IMAGES Brazil was conducted in two sites in the metropolitan area of Rio de Janeiro: Mar, a lowincome community and Vila Valquiere, a middle-income community. The sample size was calculated based on data from the Census 2000 IBGE. A stratified proportional household sample was used in each of communities with a sampling error of 5 percent. The survey instrument was administered by fieldworkers who conducted face-to-face interviews with a total of 750 men and 448 women ages 18-59 between both communities. The refusal rate was 10 percent in Mar and 45 percent in Vila Valquiere. The data collection took nearly four months. There were more difficulties in the middleincome community due to high refusal rates. The mens questionnaire contained 343 questions with a mean time of 43 minutes of application while the women questionnaire contained 255 questions and a mean time of 32 minutes of application. All respondents had to sign an informed consent form before application of the questionnaire. Eleven male and seven female interviewers were trained for 8 hours. Along with training on application of the survey instrument, they also received preparation on research techniques, data collection and ethical issues. Chile The IMAGES study in Chile had a sample of 426 women and 1192 men age 18 to 59. The sample of men was conducted in 3 major urban centers: Gran Santiago, Valparaiso and Gran Gran Concepcin, with a combined sampling error of 2.8 percent. The sample of women was held in Gran Santiago with a sampling error of 5 percent. The survey was conducted among people living in private households in urban centers. The sample was geographically stratified, with clusters at three stages: First Stage Unit: Census block or manzana, chosen based on its size. Second Stage Unit: Private, permanently inhabited homes were selected with equal probability within each block. Third Stage Unit: A person 18 to 59 years old was selected with equal probability in each house by a random selection table (Kish Table). The survey was anonymous. Informed consent was obtained verbally because in Chile there is a high fear of fraud from signing documents. Additionally, each respondent was given a written copy of the consent form explaining the objectives of the study and their rights as participants, including contact details of the researchers responsible. The questionnaire was performed in all cases as a one-on-one interview, with the option to selfadminister a later section on violence and other sensitive topics. After the interview each questionnaire was kept in a sealed envelope. Each participant was given a handout with information on health networks to contact or visit on issues such as GBV, HIV / AIDS, drugs, child care, and health. Additionally, women were given a booklet on the GBV National Service for Women. Men were interviewed by male interviewers and women by female interviewers. The interviewers were trained and monitored continuously. Ethical procedures of this study were approved by the Chilean Institute ICMER Reproductive Medicine.
70
ANNEXES
Croatia IMAGES survey data for Croatia were collected from 1501 men and 506 women aged 18-69. Respondents were selected from one urban area (Zagreb) and two rural counties in the Eastern part of the country (Osjecko-baranjska and Vukovarsko-srijemska counties). In these two counties, respondents from around 100 towns and villages took part in the survey. field managers and data collectors reported that the response rate was high.
India IMAGES in India was conducted in Delhi (urban center, North India) and Vijayawada (rural location, Tamil Nadu state of South India). In Delhi 1037 male and 313 women were interviewed for the quantitative survey. Three districts (Central Delhi, East Delhi and South The sample was stratified by age and place of living. Delhi) were selected randomly and then using Probability Proportion to Size (PPS) sampling procedure, 15 Random sampling was used to select a sufficient urban wards were selected; further two Census Enunumber of respondents from each stratum. The percentages of men and women from particular age meration Blocks (CEB) were selected randomly from the selected urban wards. A complete house listing was groups were calculated according to the 2001 Cendone in the CEBs and list of all the eligible respondents sus. Self-administered questionnaires were used. were marked and then using systematic random samAround 70 trained interviewers were responsible for pling, around 30 households were selected for identifying respondents according to their sample tem- interviewing eligible men and 10 household were plates. Sample templates included the number of about selected for interviewing eligible women respondents. 30 respondents with two criteria: age and place of living. In Vijayawada, 497 men and 208 women were interInformed consent was obtained but respondents could viewed using the same procedure. Five urban wards at any point decide to stop filling in the questionnaire. were selected using PPS sampling and three CEB Respondents had privacy while filling in the question- were selected randomly from each selected urban ward. A complete house listing was done in the naire but interviewers were accessible for additional selected CEB and list of all the eligible respondents clarification. After completing the questionnaires, were marked and then using systematic random respondents placed them into sealed envelopes and sampling, around 34 households were selected for interviewers pasted two labels on the envelope (one consisting of the info on the process of administering interviewing eligible men and 14 household were and the other to prevent the opening of the envelope) selected for interviewing eligible women In Croatia a refusal rate was not calculated. However respondents in each CEB.
Croatia Survey Sites Zagreb
Sex of the Respondents
Number of Respondents (Response Rate) 789 (52.6%) 259 (51.2%) 692 (46.1%) 242 (47.8%)
Men* Women**
Osjecko-baranjska and Vukovarsko-srijemska County
Men Women
* Men: 1.3% is missing (n=20) ** Women: 1% is missing (n=5)
71
Only one interview was conducted from selected households, ensuring that no male and female respondents came from the same household. Kish tables were used if more than one eligible respondent lived in the same household. The refusal rate was quite high in both locations.
And to ensure the required number of men and women are interviewed, a circular systematic random sampling was applied in all the survey sites, separately for men and women. This resulted in visiting more households than originally planned for the interviews.
India Survey Sites Delhi
Sex of the Respondents Men Women
Total number of HH visited for Interview 1917 655 933 496
Success 1037 313 497 208
Response Rate 54% 48% 53% 42%
Vijayawada
Men Women
Mexico In IMAGES in Mexico, the study population is aged 18 to 59 years for both men and women. Only evermarried women were interviewed. The survey was conducted in Monterrey (one of Mexicos three major metropolitan areas), Queretaro and Xalapa (two intermediate-sized cities). The design of the sample through Primary Sampling Units was contracted to the National Institute of Statistics, Geography and Informatics (INEGI). The sample design was designed to produce survey results generalizable to the population of the three cities. The design used a stratified cluster methodology, where the last unit of selection is a housing unit that contains at least one eligible respondent aged 18 to 59. The sampling frame used is INEGIs National Housing Framework 2002, which draws on the geographic and demographic information obtained by the 2000 Census of Population and Housing. Primary Sampling Units were identified in designing the IMAGES sampling frame, from which were selected housing units based on other INEGI
survey samples. The primary sampling units consist of groups of dwellings, and in the case of localities over 100,000 inhabitants, the size of a PSU ranges from 80 to 160 inhabited households. A PSU can comprise the following characteristics: a block or manzana, the union of two or more contiguous blocks of the same BGA (Basic Geostatistics Area), the union of two or more contiguous blocks of different census tracts in the same locality, or the union of two or more contiguous blocks of different locations but the same size of locality. High levels of violence in the research sites appear to be associated with high levels of non-response that occurred when trying to conduct interviews. Given a high incidence of violence related to drug trafficking in the country, it is understandable that many people allude to a fear of abduction, fraud and extortion in refusing to participate in the survey. Researchers were able to interview 1001 men and 383 women, according to the distribution described below, comparing the data with the estimate was for each city and town.
72
ANNEXES
Sex Male
Municipality Quertaro Xalapa Monterrey
Houses 350 * 350 800 1500
PSUs 35 35 80 150 12 12 26 50 200
subtotal Women Quertaro Xalapa Monterrey subtotal Total
120 120 260 500 2000
* These sample sizes were proposed by researchers at the Colegio de Mexico, based on the resources available for the collection of information through interviews.
Distribution of Primary Sampling Units, Mexico
Mexico Location Quertaro
Sex
Number of Completed Interviews 222 127 264 84 515 172 1001 383
Men Women
Jalapa
Men Women
Monterrey
Men Women
Total
Men Women
73
Rwanda IMAGES in Rwanda was conducted among men and women between 18-59 years old at a national level (covering all provinces of the country). The study sites in districts and villages were selected by cluster sampling exercised by the National Institute of Statistics of Rwanda. The study design has been approved by the same institute. For selection of the households within the villages a multi stage sampling was applied. The sample was stratified in such a way that it would include 2/3 men and 1/3 women from each age group. All five provinces are represented, with a slight difference in numbers of completed questionnaires. The quantitative data have been collected in a survey carried out by 72 trained data collectors. The women were interviewed by female data collectors, while male data collectors conducted the interviews with men. In total we received 3612 completed questionnaires, 2301 by men and 1311 by women. All procedures on ethical considerations have been followed and all survey interviews were carried out only after receiving written or oral informed consent from the participants.
The data collectors had to be trained within six days in the application of the survey instrument on interview application as well as gender issues. The field teams managed to carry out all planned interviews as is demonstrated by the total number of completed forms.
Rwanda Province Eastern province
Districts
Bugesera, Gatsibo, Kayonza,Kirehe, Ngoma, Nyagatare,Rwamagana Gasabo, Kicukiro, Nyarugenge Burera, Gakenke, Gicumbi, Musanze, Rulindo Huye, Gisagara, Kamonyi, Muhanga, Nyamagabe, Nyaruguru, Nyanza, Ruhango Karongi, Ngororero, Nyabihu,Nyamasheke, Rubavu, Rusizi, Rutsiro
Questionaires
24.3%
Kigali
11.0%
Northern province
19.2%
Southern province
25.0%
Western province
19.8%
74
Gary Barker/ICRW
ANNEXES
Annex V: Intimate Partner Violence Prevalence, IMAGES and other Recent Studies
The table below presents the IMAGES studys intimate partner violence prevalence figures in percentages side-by-side with those of other major studies in the same countries.
Year Brazil (IMAGES) Brazil (Sao Paulo)34 Brazil (Pernambuco)35 Chile (IMAGES) Chile (Sta. Rosa)36 Croatia (IMAGES) Croatia (National Study on Family Violence)37 India (IMAGES) India (DHS)38 Mexico (IMAGES) Mexico (DHS)39 Rwanda (IMAGES) Rwanda (DHS)40 2010 2001 2001 2010 2004 2010 2003
Ever perpetrated IPV (Mens reports) 24 30 33
Ever experienced IPV (Womens reports) 28 27 34 31 25 38 29
Experienced IPV in last year (Women's reports) 8 8 13 4
2010 2006 2010 2003 2010 2005
37 17 39
31 35 31 41 30
19 21 10 6 27
75
Annex VI: Tests of Association

The tables on the following pages show the frequencies and tests of association between the four key independent variables of the present paper (age, education, GEM Scale, and work-related stress) and the dependent variables of analysis (organized by topic). Cells where percentage figures appear in red italicized print represent those associations that were statistically significant at a p < .05 level. The Total Respondents column presents the total number of respondents in any particular independent variable category (age, education, GEM Scale, work-related stress) among respondents who answered that particular module of questions. The Total Respondents column refers to the following populations in each set of tables:
Topic Relationship Dynamics Parenting Physical and Sexual Health Mental Health Criminal Behaviors IPV and Sexual Violence Transactional Sex
Population represented by Total Respondents Only men with a regular or stable partner Only men with biological children living in their household All men* All men All men Only men who have ever had a female partner All men
* Under Physical and Sexual Health, two variables have exceptions: Ever sought a prostate exam was only asked of men age 40 or older, and Ever participated in an abortion decision was only asked of men who reported that their partners had terminated a pregnancy.
76
David Snyder/ICRW
For all variables included, he only difference between the numbers in the Total Respondents column and the actual n for any specific test is the negligible number of dont know or not available responses. These responses were coded as missing and omitted from the analyses underlying the figures presented in the tables. For the sake of table concision and readability, precise n values for every independent variable/dependent variable pairing have been omitted.
ANNEXES
Relationship Dynamics
Brazil
Age
18-24 25-34 35-49 50-59
Couple Communication 86 86 85 89 92 83 90 90 87 84 87 83 88 86 Couple Communication 92 90 87 84 70 89 87 89 50 86 90 87 88 88
Participation in Domestic Duties 59 55 65 60 38 56 64 77 46 53 65 58 60 60 Participation in Domestic Duties 62 69 51 37 50 47 45 65 19 36 61 56 49 52
Sexual Satisfaction 96 93 91 91 77 90 97 96 80 91 95 93 91 92 Sexual Satisfaction 88 94 85 76 88 83 84 88 50 83 88 88 85 86
Overall Relationship Satisfaction 93 88 87 93 85 88 90 100 88 87 91 93 89 90 Overall Relationship Satisfaction 97 94 92 94 88 92 95 93 95 96 92 96 92 94
Total Respondents
106 139 184 108 13 294 178 51 40 174 315 274 149
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary
GEM
Work Stress
No Yes
Total
Chile
Age
18-24 25-34 35-49 50-59
Total Respondents
107 302 324 185 10 99 397 312 24 227 568 312 440
Education
GEM
Work Stress
No Yes
Total
77
Croatia
Age
18-24 25-34 35-49 50-59
Couple Communication 99 96 91 90
Participation in Domestic Duties 83 75 57 55
Sexual Satisfaction 93 85 72 68
Overall Relationship Satisfaction 79 95 90 95
Total Respondents
133 333 322 255 0
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 72 94 94 70 88 95 96 93 93 Couple Communication 69 68 63 72 65 62 68 67 73 65 64 63 70 67 45 61 64 28 49 65 67 60 61 Participation in Domestic Duties 23 16 15 17 13 16 18 16 15 17 14 14 17 16 66 76 80 61 67 79 84 72 77 Sexual Satisfaction 98 98 98 97 98 95 99 97 98 98 98 99 98 98 83 95 91 95 92 93 96 91 93 Overall Relationship Satisfaction 100 98 97 98 99 94 99 98 99 97 98 99 97 98 Total Respondents 34 625 418 23 162 847 355 412
GEM
Work Stress
No Yes
Total
India
Age
18-24 25-34 35-49 50-59
52 333 425 122 133 98 342 359 221 570 141 360 526
Education
GEM
Work Stress
No Yes
Total 78
ANNEXES
Mexico
Age
18-24 25-34 35-49 50-59
Couple Communication 94 89 91 83 70 77 86 94 71 85 91 91 89 89 Couple Communication
Participation in Domestic Duties 60 56 56 45 50 47 48 60 35 39 60 58 52 54 Participation in Domestic Duties 62 57 48 51 49 52 58 58 43 47 65
Sexual Satisfaction 100 98 95 88 100 86 95 97 86 91 96 93 95 95 Sexual Satisfaction 86 87 88 83 85 87 81 94 83 85 90
Overall Relationship Satisfaction 89 93 94 95 90 96 93 94 100 94 93 94 94 94 Overall Relationship Satisfaction
Total Respondents
151 196 273 121 11 90 186 454 24 161 556 80 527
Education
GEM
Work Stress
No Yes
Total
Rwanda
Age
18-24 25-34 35-49 50-59
Total Respondents
98 644 730 443 353 1248 229 85 306 999 555
Education
GEM
Work Stress
No Yes
Total
52
86
79
Parenting
Brazil
Age
18-24 25-34 35-49 50-59
Presence at last birth 41 40 51 48 33 38 55 83 23 41 53 50 39 46 Presence at last birth 95 85 70 61 44 53 71 82 44 69 75 79 68 72
Accompaniment to prenatal visit 83 83 79 70 60 71 90 97 58 68 87 81 78 78 Accompaniment to prenatal visit 95 90 87 78 67 70 86 92 27 83 89 85 87 86
Took leave at birth of last child 55 66 67 51 75 57 67 71 56 58 66 67 53 62 Took leave at birth of last child 73 34 23 10 29 17 20 31 7 19 26 24 23 23
Participation in daily care of child 42 33 43 38 13 36 42 67 25 34 42 34 38 39 Participation in daily care of child 40 35 36 37 13 35 33 44 8 31 40 37 35 36
Total Respondents
47 118 182 122 18 285 126 39 38 164 255 236 133
Education
GEM
Work Stress
No Yes
Total
Chile
Age
18-24 25-34 35-49 50-59
Total Respondents
40 131 309 197 10 93 356 217 21 204 452 260 385
Education
GEM
Work Stress
No Yes
Total 80
ANNEXES
Croatia
Age
18-24 25-34 35-49 50-59
Presence at last birth 33 66 49 32
Accompaniment to prenatal visit 100 96 92 89
Took leave at birth of last child 33 57 39 37
Participation in daily care of child 67 55 61 67
Total Respondents
3 92 259 248 0
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 44 39 55 27 43 47 55 43 46 Presence at last birth 37 76 73 76 69 67 71 76 65 74 75 76 74 72 83 89 96 69 84 94 96 92 92 Accompaniment to prenatal visit 89 92 91 90 90 87 91 93 91 91 93 85 96 91 50 42 38 8 29 43 44 40 40 Took leave at birth of last child 75 72 66 67 59 69 65 75 61 71 69 78 62 69 45 61 66 36 45 67 70 55 63 Participation in daily care of child 43 37 37 36 39 32 36 39 38 35 45 35 38 37 Total Respondents 27 370 237 17 107 478 196 263
GEM
Work Stress
No Yes
Total
India
Age
18-24 25-34 35-49 50-59
22 242 414 131 127 87 302 293 194 489 126 317 449
Education
GEM
Work Stress
No Yes
Total
81
Mexico
Age
18-24 25-34 35-49 50-59
Presence at last birth 28 29 29 18 50 18 18 33 31 27 26 25 27 27 Presence at last birth 45 54 50 49 40 53 53 63 58 48 52
Accompaniment to prenatal visit 96 92 93 89 100 85 89 95 75 87 94 92 92 92 Accompaniment to prenatal visit
Took leave at birth of last child 80 71 62 68 60 68 67 65 43 63 68 67 66 66 Took leave at birth of last child 52 49 46 51 55 46 52 55 41 46 56
Participation in daily care of child 52 42 44 54 50 50 45 45 44 37 49 47 43 46 Participation in daily care of child
Total Respondents
33 132 259 125 13 85 167 284 22 132 395 61 428
Education
GEM
Work Stress
No Yes
Total
Rwanda
Age
18-24 25-34 35-49 50-59
Total Respondents
52 595 729 463 347 1202 214 77 291 969 534
Education
GEM
Work Stress
No Yes
Total 82
51
48
ANNEXES
Physical and Sexual Health
Brazil
Age
18-24 25-34 35-49 50-59
Binge drinking
Prostate Exam (among those age 40+)
HIV exam
Participated in abortion decision (among those whose partners terminated a pregnancy)
Total Respondents
67 72 68 69 69 69 67 78 64 69 70 69 73 69 Binge drinking 21 44 50 28 38 41 37 29 35 25 39 33
26 34 41 38 24 32 39 39 28 31 38 35 39 35 HIV exam
12 31 30 50 0 27 37 50 18 29 30 29 39 30
176 187 236 151 21 401 238 89 65 243 427 366 190
Education
GEM
Work Stress
No Yes
Total
Chile
Age
18-24 25-34 35-49 50-59
Total Respondents
53 44 32 24 33 43 37 41 69 47 35 35 38 39
16 36 33 29 35 22 26 33 21 24 31 33 28 29
67 47 71 76
289 283 391 229 17
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 50 68 66 67 78 62 66 67 65 127 566 481 35 321 835 432 578
GEM
Work Stress
No Yes
Total
83
Croatia
Age
18-24 25-34 35-49 50-59
Binge drinking
HIV exam
Total Respondents
64 48 33 23 100 42 44 38 56 49 40 38 40 42 Binge drinking 30 35 53 24 37 44 49 39 42

8 13 33 49 12 10 0 4 9 16 5 3 13 14 12 11 HIV exam
50 67 72 69
277 498 389 290 1
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 63 75 62 75 69 70 64 74 71
55 902 537 39 220 1161 475 534
GEM
Work Stress
No Yes
Total
India
Age
18-24 25-34 35-49 50-59
Total Respondents
16 23 27 21 36 12 18 25 25 21 28 14 31 23 8 38 25 19 17 26 28 20 16 10 21 22
8 17 14 15 12 16 13 13 20 12 6 12 15 13
83 78 41
469 494 452 137
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 67 50 55 62 80 50 180 133 533 706 366 923 263 60 57 59 466 762
GEM
Work Stress
No Yes
Total 84
ANNEXES
Mexico
Age
18-24 25-34 35-49 50-59
Binge drinking
HIV exam
Total Respondents
33 41 33 26 21 37 37 33 50 39 32 39 35 34 Binge drinking 16 25 0 12 12 26 13 20 20 5 21 20
13 36 29 29 14 17 15 33 16 15 31 20 30 27 HIV exam
33 19 15 12 0 14 21 18 40 27 14 33 17 18
283 250 323 146 14 116 245 627 37 220 745 96 674
Education
GEM
Work Stress
No Yes
Total
Rwanda
Age
18-24 25-34 35-49 50-59
Total Respondents
73 94 90 78 81 87 91 91 85 87 89
303 760 768 470 406 1481 312 102 363 1160 647
Education
GEM
Work Stress
No Yes
Total
87
85
Mental Health
Brazil
Age
18-24 25-34 35-49 50-59
Experienced depression in last month 13 13 14 23 19 13 18 18 11 19 13 9 24 15 Experienced depression in last month 34 36 41 39 41 47 39 33 43 39 37 27 46 38
Suicidal thoughts in last month 5 3 3 5 5 3 4 4 3 5 4 2 7 4 Suicidal thoughts in last month 8 12 8 7 12 17 9 7 11 10 8 5 10 9
Seeks help when sad, disappointed, frustrated 55 52 54 44 38 52 55 48 43 53 52 49 53 52 Seeks help when sad, disappointed, frustrated 66 60 51 43 35 40 53 63 26 49 59 63 48 55
Total Respondents
176 187 236 151 21 401 238 89 65 243 427 366 190
Education
GEM
Work Stress
No Yes
Total
Chile
Age
18-24 25-34 35-49 50-59
Total Respondents
289 283 391 229 17 127 566 481 35 321 835 432 578
Education
GEM
Work Stress
No Yes
Total 86
ANNEXES
Croatia
Age
18-24 25-34 35-49 50-59
Experienced depression in last month 62 66 67 66 100 66 68 60 81 71 64 51 79 65 Experienced depression in last month 56 54 53 45 47 58 58 51 25 59 73 38 65 54
Suicidal thoughts in last month 17 15 17 9 100 30 16 12 31 20 13 10 19 15 Suicidal thoughts in last month 7 5 7 9 4 8 8 6 4 8 6 4 9 7
Seeks help when sad, disappointed, frustrated 73 70 66 61
Total Respondents
277 498 389 290 1
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 56 64 72 42 62 70 71 67 67 Seeks help when sad, disappointed, frustrated 12 11 10 11 13 7 12 11 15 11 5 12 9 11 Total Respondents 55 902 537 39 220 1161 475 534
GEM
Work Stress
No Yes
Total
India
Age
18-24 25-34 35-49 50-59
469 494 452 137 180 133 533 706 366 923 263 466 762
Education
GEM
Work Stress
No Yes
Total
87
Mexico
Age
18-24 25-34 35-49 50-59
Experienced depression in last month 27 30 27 40 36 33 29 29 35 35 28 44 27 30
Suicidal thoughts in last month 2 4 3 4 7 2 4 3 0 4 3 7 2 3
Seeks help when sad, disappointed, frustrated 47 52 49 41 29 39 49 50 28 40 51 53 47 48
Total Respondents
283 250 323 146 14 116 245 627 37 220 745 96 674
Education
GEM
Work Stress
No Yes
Total
88
ANNEXES
Criminal Violence
Brazil
Age
18-24 25-34 35-49 50-59
Robbery
Fights
Owns Firearm
Arrested
Prison
Total Respondents 176 187 236 151 21 401 238 89 65 243 427 366 190
13 10 12 5 15 12 10 3 11 11 10 9 10 10 Robbery
25 20 22 19 30 22 25 9 16 24 21 21 27 22 Fights
4 1 4 3 0 3 2 2 2 0 4 3 2 3 Owns Firearm
13 13 14 13 15 17 10 6 13 18 11 13 14 13 Arrested
5 2 2 2 0 4 2 0 5 4 2 3 3 3 Prison
Education
GEM
Work Stress
No Yes
Total
Chile
Age
18-24 25-34 35-49 50-59
31 27 17 13 24 23 24 19 26 24 21 20 21 22
24 20 10 8 12 20 17 12 29 16 15 13 17 16
5 9 10 15 24 8 8 12 15 15 8 10 9 10
29 29 31 26 41 41 31 23 44 26 30 25 32 29
1 5 5 5 6 12 4 3 9 5 4 3 6 4
Education
GEM
Work Stress
No Yes
Total
89
Croatia
Age
18-24 25-34 35-49 50-59
Robbery
Fights
Owns Firearm
Arrested
Prison
44 46 31 21 100 46 40 29 54 38 35 31 40 36 Robbery
17 18 18 18 0 40 19 14 43 30 15 14 21 18 Fights
3 7 14 20 0 12 11 11 19 15 10 10 11 11 Owns Firearm
14 27 14 15 0 24 21 14 22 19 18 17 20 18 Arrested
3 7 6 8 0 12 6 5 3 7 6 6 6 6 Prison
Education
GEM
Work Stress
No Yes
Total
India
Age
18-24 25-34 35-49 50-59
4 6 2 4 6 2 5 4 4 5 2 1 6 4
5 9 6 5 5 11 6 7 7 8 1 7 6 7
1 1 2 0 1 0 2 1 2 1 1 1 2 1
3 9 6 9 8 8 6 5 9 6 3 6 7 6
0 1 1 3 1 2 1 1 2 1 0 0 1 1
Education
GEM
Work Stress
No Yes
Total
90
ANNEXES
Mexico
Age
18-24 25-34 35-49 50-59
Robbery
Fights
Owns Firearm
Arrested
Prison
4 3 3 3 0 0 6 3 16 4 2 7 2 3 Robbery
17 12 12 5 0 9 16 12 22 12 12 13 11 13 Fights
3 3 3 5 0 3 3 4 8 2 4 7 3 3 Owns Firearm
28 26 27 31 29 37 34 23 49 34 25 23 28 28 Arrested
2 2 4 6 21 8 4 2 3 4 3 1 4 3 Prison
Education
GEM
Work Stress
No Yes
Total
Rwanda
Age
18-24 25-34 35-49 50-59
Total Respondents 303 760 768 470 406 1481 312 102 363 1160 647
18 16 11 6 9 13 16 15 14 13 11
4 7 5 2 2 4 11 8 5 6 3
2 5 3 5 4 3 5 8 6 4 3
15 21 28 30 31 23 22 24 21 26 26
8 12 21 22 20 16 13 14 18 16 17
Education
GEM
Work Stress
No Yes
Total
13
25
16
91
IPV and Sexual Violence
Brazil
Age
18-24 25-34 35-49 50-59
Physical violence against Physical violence against partner ever partner in the last year
Sexual violence ever
Sexual violence in a group ever
Total Respondents
25 19 31 19 24 29 23 9 33 26 22 23 29 24
18 9 10 4 14 12 11 2 20 13 8 8 14 11
4 2 2 1 0 3 2 0 0 4 1 2 2 2
2 1 1 0 0 2 1 0 0 1 1 1 2 1
173 185 231 149 21 395 234 87 64 239 421 361 188
Education
GEM
Work Stress
No Yes
Total
Chile
Age
18-24 25-34 35-49 50-59
Total Respondents
17 30 35 35 40 41 30 26 64 32 27 27 35 29
10 9 9 10 6 8 9 10 27 11 8 10 9 9
2 2 1 1 0 2 1 1 3 2 1 2 1 2
272 271 381 223 15 121 544 466 33 309 805 420 559
Education
GEM
Work Stress
No Yes
Total 92
ANNEXES
Croatia
Age
18-24 25-34 35-49 50-59
Total Respondents
18 29 39 47 0 39 34 30 63 43 30 28 36 33
11 9 8 6
2 2 1 0
259 470 352 247 1
Education
No Formal Education Up to Primary Class V Primary to Senior Secondary Beyond Senior Secondary 14 9 7 63 17 6 6 11 9
0 2 1 16 3 1 1 2 2
41 826 496 35 200 1067 436 486
GEM
Work Stress
No Yes
Total
India
Age
18-24 25-34 35-49 50-59
Total Respondents
16 34 40 49 43 52 38 30 39 39 29 30 43 37
7 19 19 21 27 29 20 12 20 19 14 12 24 19
11 30 34 17 28 29 25 22 27 26 14 30 27 24
0 1 0 1 0 1 1 0 1 1 0 0 1 1
55 327 420 127 135 96 343 355 214 575 140 356 529
Education
GEM
Work Stress
No Yes
Total
93
Mexico
Age
18-24 25-34 35-49 50-59
Physical Violence Against Physical violence against partner ever partner in the last year
Total Respondents
14 18 18 22 21 18 22 15 49 21 15 28 18 17
Physical Violence Against Physical violence against partner ever partner in the last year
6 5 3 1 0 3 8 3 14 8 3 6 4 4
1 0 1 0 0 1 2 0 3 1 0 0 1 1
273 249 323 144 14 115 242 618 37 216 736 96 670
Education
GEM
Work Stress
No Yes
Total
Rwanda
Age
18-24 25-34 35-49 50-59
Total Respondents
28 38 38 42 40 38 43 30 45 40 34
3 2 3 1 2 2 4 0 2 3 1
4 11 10 8 10 8 13 9 12 11 6
1 1 1 1 2 1 1 0 1 1 1
61 553 669 426 301 1137 198 73 264 889 516
Education
GEM
Work Stress
No Yes
Total 94
39
ANNEXES
Transactional Sex
Brazil
Age
18-24 25-34 35-49 50-59
Sex with any sex worker
Sex with female sex worker
Sex with male sex worker
Sex with transgender sex worker
Total Respondents
50 60 56 58
79 53 56 65
47 58 54 56
74 50 54 65
1 0 1 0 0 0 1 0 0 0 1 0 1 0
1 2 2 1 5 2 2 0 2 0 2 2 2 2
176 187 236 151 21 401 238 89 65 243 427 366 190
Education
GEM
51 53 59 57 59 56
Sex with any sex worker 12 20 26 32
49 53 56 55 56 54
Sex with female sex worker 12 19 25 32
Work Stress
No Yes
Total
Chile
Age
18-24 25-34 35-49 50-59
Total Respondents
0 0 1 0 0 2 1 0 0 0 1 0 1 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0
289 283 391 229 17 127 566 481 35 321 835 432 578
Education
38 25 22 22
48 26 20
38 23 21 22
48 25 19
GEM
Work Stress
No Yes
26 22 22
25 21 22
Total
95
Croatia
Age
18-24 25-34 35-49 50-59
Sex with any sex worker 6 12 13 15 100 20 12 10 42 20 9 9 14
Sex with female sex worker 6 11 13 15 100 20 12 10 42 20 9 9 14
Total Respondents
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
277 498 389 290 1 55 902 537 39 220 1161 475 534
Education
GEM
Work Stress
No Yes
Total
12
Sex with any sex worker
11
Sex with female sex worker 19 25 28 27 34 35 24 19 24 26 16 21 28
India
Age
18-24 25-34 35-49 50-59
Total Respondents
23 28 30 30
35 38 27 23 24 30 23 24 31
3 2 1 1 1 1 2 2
0 2 5
1 1 2 1 0 2 1 2 0 2 2 2 1 1
469 494 452 137 180 133 533 706 366 923 263 466 762
Education
GEM
Work Stress
No Yes
1 2 2
Total 96
27
24
ANNEXES
Mexico
Age
18-24 25-34 35-49 50-59
Sex with female sex worker 8 17 22 26
Total Respondents
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
283 250 323 146 14 116 245 627 37 220 745 96 674
Education
7 16 18 18
35 18 17
7 16 18 18
35 18 17
GEM
Work Stress
No Yes
21 19 18
21 19 18
Sex with female sex worker 9 15 16 17 12 15 19 22 18 16 11
Total
Rwanda
Age
18-24 25-34 35-49 50-59
Total Respondents
0 0 0 0 0 0 0 0 0 0 0
1 1 1 0 1 1 0 0 1 1 1
303 760 768 470 406 1481 312 102 363 1160 647
Education
13 15 19 22
19 17 12
GEM
Work Stress
No Yes
Total
16
15
97
Notes
The IMAGES survey will be carried out in additional countries in 2011 and thereafter. 2 The GEM Scale is a collection of attitude questions widely used in diverse settings and consistently show high rates of internal reliability. 3 U.N. Division of Advancement of Women, The Agreed Conclusions of the 48th session of the Commission on the Status of Women available at: http://www.un.org/womenwatch/ daw/csw/csw48/ac-men-auv.pdf 4 An overview of these policies is presented in the final section of this chapter. 5 For more information about MenEngage, see www.menengage.org or www.engagingmen.net. 6 As of 2010-2011, additional countries will be carrying out an adapted version of IMAGES in Asia and the Pacific region as part of Partners for Prevention the U.N. Joint Programme for Ending Violence Against Women in Asia and the Pacific, based in the U.N. regional office Bangkok, Thailand, and in other parts of the world with other partners. 7 These items include the Questionnaire on Gender Equality and Quality of Life implemented by the Norwegian Ministry of Gender Equality and Children Affairs, the WHO multicountry study on violence against women, the Gender Equitable Men (GEM) Scale, developed by Population Council and Promundo, and surveys on sexual violence and physical violence against women carried out the Medical Research Council in South Africa. 8 World Health Organization and PATH (2005). Researching Violence Against Women: A Practical Guide for Researchers and Activists. Available online at: http://www.who.int/ gender/documents/women_and_girls/9241546476/en/ index.html 9 A more detailed description of the sampling methodology in each setting is included in Annex IV. 10 This study is available for download at: http://www.nikk.no/ 11 This study is available for download at: http://www.mrc.ac.za/ 12 The full IMAGES questionnaires for women and men are available online at: www.icrw.org and www.promundo.org. 13 This section makes ample use of relevant WHO and other U.N. instruments, including relevant items from the Demographic and Health Surveys. The IMAGES questionnaire includes relatively few questions about contraceptive use, condom use, STIs and HIV precisely because these issues have frequently been included in other studies. 14 Specific questions about work-related stress asked of all men (regardless of current employment status) were: I am frequently stressed or depressed because of not having enough work and I am frequently stressed or depressed because of not having enough income. Possible responses were: yes, partly, no, dont know. Throughout this report, the independent variable work-related stress refers to mens responses to either or both of these two items. Unemployed men were also asked specific questions about feeling ashamed to face their family, having considered leaving their family and or whether they drink or stay away from home when they cannot find work. 15 The GEM Scale was originally developed by the Population Council and Promundo with young men aged 15-24 years (Pulerwitz and Barker, 2008). The original scale includes 17
1
attitudinal statements about different dimension of mens gender attitudes. The scale has since been validated in household research in more than 15 countries. For IMAGES, the GEM Scale was slightly adapted with additional questions appropriate for adult men. However, care was taken that each country should have at least 15 common GEMS items covering the same range of issues from the original scale: sexuality, violence, household tasks, homophobia and male/female roles. 16 Following the original GEM Scale process, principal component factor analysis was carried out to create a final GEM Scale for each country (with 11-15 items included, depending on the country). Internal consistency was assessed using reliability analysis with Cronbach Alpha scores ranging from 0.72 to 0.83, which is considered a high degree of reliability. Please see Annex III for more information. 17 Men were asked about each of these activities as separate questions. 18 Any man who specifically said that the birth happened in the home was excluded from this analysis. 19 Apart from the obvious gender equality issues involved, numerous studies have affirmed the importance of such practices for men themselves. A study in the United Kingdom found that participating in prenatal ultrasound examinations was important for the men, as it helped them to visualize the baby and realize their transition to fatherhood (Draper, 2002), and a study from Sweden concludes that that many men, in the context of Sweden, experience the ultrasound as confirmation of a new life and therefore an important milestone in developing a paternal identity (Ekelin, et al., 2004). 20 In India and other parts of Asia, ultrasound is sometimes used for sex selective abortion (of girl fetuses) and thus has obviously different social meanings depending on cultural context, mens attitudes and family decision-making processes. 21 Percentage of men with children who, when asked about the daily care of children, said Always me, usually me or equally or done together. 22 As mentioned in the introduction, one component of the Men and Gender Equality Policy project is a multi-country qualitative study on men in care-giving roles (either in their domestic/personal lives or as a profession). These results, which aim to provide greater understanding on men who are the chief caregivers in their families, will complement the findings from IMAGES. The final qualitative report will be published later in 2011. 23 IMAGES did not use any of the standard instruments on depression, which provide a list of symptoms. Instead, it simply asked about experiencing depression in the last month (never, seldom, sometimes, often). 24 In Annex V, we provide comparisons between the WHO data and IMAGES data. 25 In the questionnaire, physical violence includes the following acts: slapping, throwing objects, pushing, shoving, hitting, kicking, dragging, beating, choking, burning, threatening or using of weapons. 26 In Brazil, India and Rwanda, both women and men were asked if any of the forms of IPV experienced or used had been perpetrated during the last 12 months.
98
NOTES
27
The levels of IPV found in Mexico from the male questionnaire are clearly underestimated. Different national surveys in this country show higher levels similar to those found from the IMAGES females questionnaire (higher than 30 percent). Several methodological issues could be related to this finding, particularly the fact that some women interviewers were employed to interview men (although no men interviewed women). 28 These rates include mens self-reports of sexual violence against other boys or men, but these were very few in number (six cases in total of all the men who responded to the questions). 29 Sexual violence questions included forced sex (How many times have you had sex with a woman or girl when she didnt consent to sex or after you forced her?); having sex with a women or girl if she had had too much to drink (How many times have you had sex with a woman or girl when she was too drunk to say whether she wanted it or not?); specific questions about forced sex with either a current wife or girlfriend, an ex-girlfriend or ex-wife, forced sex with any other partner; forced sex with a boy or man; and gang rape with other men or boys against a woman or girl. IMAGES items on sexual violence were informed by the work of the Medical Research Council of South Africa, particularly the recent study (Jewkes, et al 2009) on men, health and violence, which takes into account the most recent research on factors associated with mens use of sexual violence as well as methodological advances in asking sensitive questions 30 Questions about sexual violence were not included in the womens questionnaire, thus we cannot compare mens reports to womens reports of sexual violence. 31 The majority of respondents in the case of Brazil were from a low income urban setting (favela) in Rio de Janeiro, were previous research has found similarly high rates of incarceration and conflict with the law.
32
Though not reported here, IMAGES found significant percentages of men in the survey sites had witnessed violence by a man against their mother and/or experienced physical violence from a parent while growing up. A more detailed analysis of these childhood experiences from IMAGES will be published in a forthcoming report. 33 Sex work is legal in Brazil (except with minors). Moreover, having had sex with a sex worker is seen as a legitimate sexual practice among many Brazilian men, so it is not surprising that mens reported rates of having paid for sex were so high. 34 Garca-Moreno, C., Janse, H., Ellsberg, M., Heise, L., Watts, C. (2005). WHO Multi-country Study on Womens Health and Domestic Violence against Women. Initial results on prevalence, health outcomes and womens responses. Geneva, Switzerland: World Health Organization. 25 Same as previous note 36 Ellsberg, M., Heise, L. (2005). Researching Violence Against Women: a practical guide for researchers and activists. Washington, D.C.: WHO, PATH. 37 Bego, A. et al. (2007). National Study on Family Violence against Women in the Republic of Croatia. Croatia: Center for Women War Victims-Rosa, Autonomous Womens House. 38 International Institute for Population Sciences and Macro International (2007). National Family Health Survey 20052006. India. Vol. I. Mumbai: IIPS. 39 Instituto Nacional de Estadstica, Geografa e Informtica (INEGI). (2004). Encuesta Nacional sobre la Dinmica de las Relaciones en los Hogares 2003 (ENDIREH). Tabulados Bsicos. Aguascalientes, Mxico: INEGI. 40 Institut Nacional de la Statistique du Rwanda (INSR) and Macro International (2006). Rwanda Demographic and Health Survey 2005. Maryland, USA: INSR.
99
International Center for Research on Women (ICRW)

1120 20th St. N.W. Suite 500 North Washington, D.C. 20036 USA Phone: +1-202-797-0007 Fax: +1-202-797-0020 E-mail: info@icrw.org Web: www.icrw.org
Asia Regional Office

C 139 Defence Colony New Delhi 110024 India Phone: +91-11-2465-4216 Fax: +91-11-2463-5142 E-mail: info.india@icrw.org
Instituto Promundo
Rua Mxico, 31 /1502 - Centro Rio de Janeiro - RJ - Brazil Cep. 20031-904 Phone/Fax: +55-21-2215-5216 E-mail: promundo@promundo.org.br Web: www.promundo.org.br

Evolving Men IMAGES

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Evolving Men IMAGES

Uploaded by

Copyright:

Available Formats

Evolving Men

About the Men and Gender Equality Policy Project

PART II: Results

PART III: Conclusions References Annexes

Tables and Figures

Marie Swartz/Influential Men

53 35 Mens Attitudes about Male Sex Workers Under 18 54 55 56 57 58

Jon Spaull/Influential Men

Ping-hang Chen/Influential Men

In-country research partner

El Colegio de Mxico, Rwanda Mens Mxico, D.F Resource Centre (RWAMREC)

Questionnaire Application Process

Paper Survey (mix of self administered and interviewer administered)

Paper Survey (mix of self administered and interviewer administered)

Paper Survey (interviewer administered)

Table 1 - IMAGES Data Collection Details

Conceptual Framework, Topics and Analytical Strategy

David Isaksson/Influential Men

PART II: Results

Table 2 - Socio-demographic Characteristics, Men

Table 3 - Socio-demographic Characteristics, Women

Brazil 34 Chile 57 Croatia 53 India 62 Mexico 88

% of men reporting stress/depression related to work or income

Figure 1: Mens Reports of Work-related Stress

Brazil GEM Scale Items by Domains Gender

Brazil 21 43 Chile 55 44 Croatia 63 63 India

% of men who agree

Figure 2: Mens Attitudes about Homosexuality

Richard Lewisohn/Influential Men

86 88 Chile 88 90 Croatia 93 91 India 67 70 Mexico 89 88

Mens Practices and Lives

Brazil 26 Chile 52 35 Croatia

61 47 India 16 18 Mexico 54 47 Rwanda 52 23

Figure 4: Mens and Womens Reports of Mens Participation in Domestic Duties

Brazil 58 60 Chile* 43 59 Croatia* 52

66 India 13 18 Mexico* 44 69 Rwanda* 46 60

% who play an equal or greater role in domestic duties

86 75 Croatia 77 78 India 98 96 Mexico 95 86 Rwanda 86 60

% reporting sexual satisfaction with current partner

Has not talked to partner about problems in last month

Has talked to partner about problems in last month

92 93 Chile* 67 89 Croatia* 47 80 India 98 98 Mexico* 82 96

% who report sexual satisfaction

Has not talked to partner about problems in last month

Has talked to partner about problems in last month

82 Chile* 55 78 Croatia* 48 84 India 95 96 Mexico* 63 90

% who report sexual satisfaction

90 88 Chile 94 80 Croatia 93 80 India 98 97 Mexico 94 88

% reporting satisfaction with current division of duties

% of men at specified location for birth of last child

Figure 10: Mens Presence During Birth of Last Child

18-24 90 25-34 62 35-49 52 50-59 31

% present in the delivery room for birth of last child

86 72 Croatia 92 75 India 91 86 Mexico 92 80

% who say yes, man accompanied

% who report type of leave

Brazil 10 Chile 36 10 Croatia

Womens Reports (of Men)

% who participate in the daily care of a child

Brazil 83 46 52 Chile 70 27 39 Croatia 76 45 52 India 46 18 19 Mexico 64 28 35

% who report doing activity several times a week or more

Figure 15: Mens Reports of Care of Children Age 4 and Younger

Brazil 72 25 42 Chile 95 91 Croatia 71 42 51 India 74 30 29 Mexico 46 15 23

Brazil 53 35 Chile* 57 34 Croatia